Datasets:
name stringlengths 34 57 | instruction stringclasses 24
values | description stringlengths 93 116 | keywords listlengths 5 5 | schema_version stringclasses 1
value | config unknown | metadata unknown | files listlengths 10 10 | location stringlengths 36 59 |
|---|---|---|---|---|---|---|---|---|
alabamamikertr/dd-s0250-arr-ex-coc | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is total FY2025 ARR excluding customers whose execut... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, arr_ex_coc). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-arr-ex-coc",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, arr_ex_coc).",
"authors": [],
"keywords": [
"... | {
"seed": 250,
"fact_id": "arr_ex_coc",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-arr-ex-coc |
alabamamikertr/dd-s0250-budget-growth-fy26 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What FY2026 revenue growth does management's budget imply... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, budget_growth_fy26). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-budget-growth-fy26",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, budget_growth_fy26).",
"authors": [],
"keyw... | {
"seed": 250,
"fact_id": "budget_growth_fy26",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-budget-growth-fy26 |
alabamamikertr/dd-s0250-cim-revenue-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does FY2025 revenue stated in the CIM e... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, cim_revenue_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"reconciliation"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-cim-revenue-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, cim_revenue_gap).",
"authors": [],
"keywords":... | {
"seed": 250,
"fact_id": "cim_revenue_gap",
"tier": 2,
"question_type": "reconciliation",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-cim-revenue-gap |
alabamamikertr/dd-s0250-coc-customers | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Which of the company's customers have a right under their... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, coc_customers). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"red_flag"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-coc-customers",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, coc_customers).",
"authors": [],
"keywords": [
... | {
"seed": 250,
"fact_id": "coc_customers",
"tier": 3,
"question_type": "red_flag",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-coc-customers |
alabamamikertr/dd-s0250-fy24-gross-margin | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the company's FY2024 gross margin (gross profit ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, fy24_gross_margin). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup_calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-fy24-gross-margin",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, fy24_gross_margin).",
"authors": [],
"keywor... | {
"seed": 250,
"fact_id": "fy24_gross_margin",
"tier": 1,
"question_type": "lookup_calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-fy24-gross-margin |
alabamamikertr/dd-s0250-h-cim-ebitda-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does the Adjusted EBITDA stated in the ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_cim_ebitda_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"reconciliation_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-cim-ebitda-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_cim_ebitda_gap).",
"authors": [],
"keywords... | {
"seed": 250,
"fact_id": "h_cim_ebitda_gap",
"tier": 4,
"question_type": "reconciliation_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-cim-ebitda-gap |
alabamamikertr/dd-s0250-h-coc-h1-exposure | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Customers whose executed MSA allows termination on a chan... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_coc_h1_exposure). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-coc-h1-exposure",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_coc_h1_exposure).",
"authors": [],
"keywor... | {
"seed": 250,
"fact_id": "h_coc_h1_exposure",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-coc-h1-exposure |
alabamamikertr/dd-s0250-h-diligence-ebitda | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is diligence-adjusted FY2025 EBITDA (USD): EBITDA pe... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_diligence_ebitda). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-diligence-ebitda",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_diligence_ebitda).",
"authors": [],
"keyw... | {
"seed": 250,
"fact_id": "h_diligence_ebitda",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-diligence-ebitda |
alabamamikertr/dd-s0250-h-latest-renewal | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Per the FY2026 renewal schedule, which customer (by name)... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_latest_renewal). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"pdf_table_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-latest-renewal",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_latest_renewal).",
"authors": [],
"keywords... | {
"seed": 250,
"fact_id": "h_latest_renewal",
"tier": 4,
"question_type": "pdf_table_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-latest-renewal |
alabamamikertr/dd-s0250-h-top5-enterprise-precedence | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Applying the data room's source-precedence rules for cust... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_top5_enterprise_precedence). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"calc_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-top5-enterprise-precedence",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_top5_enterprise_precedence).",
"aut... | {
"seed": 250,
"fact_id": "h_top5_enterprise_precedence",
"tier": 4,
"question_type": "calc_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-top5-enterprise-precedence |
alabamamikertr/dd-s0250-h-trap-contract-len-2023 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the average initial contract term, in months, for... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_contract_len_2023). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-trap-contract-len-2023",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_contract_len_2023).",
"authors": [... | {
"seed": 250,
"fact_id": "h_trap_contract_len_2023",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-trap-contract-len-2023 |
alabamamikertr/dd-s0250-h-trap-expansion-h2-2025-renewals | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** How much expansion ARR (USD) came from customers whose co... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_expansion_h2_2025_renewals). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-trap-expansion-h2-2025-renewals",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_expansion_h2_2025_renewals)."... | {
"seed": 250,
"fact_id": "h_trap_expansion_h2_2025_renewals",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-trap-expansion-h2-2025-renewals |
alabamamikertr/dd-s0250-h-trap-grr-fy24-cohort | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 gross revenue retention for customers fir... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_grr_fy24_cohort). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-trap-grr-fy24-cohort",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_grr_fy24_cohort).",
"authors": [],
... | {
"seed": 250,
"fact_id": "h_trap_grr_fy24_cohort",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-trap-grr-fy24-cohort |
alabamamikertr/dd-s0250-h-trap-nrr-ent-fy23-cohort | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 net revenue retention for Enterprise cust... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_nrr_ent_fy23_cohort). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-trap-nrr-ent-fy23-cohort",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_nrr_ent_fy23_cohort).",
"authors... | {
"seed": 250,
"fact_id": "h_trap_nrr_ent_fy23_cohort",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-trap-nrr-ent-fy23-cohort |
alabamamikertr/dd-s0250-h-trap-partner-churn | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** How much FY2025 ARR (USD) was lost from customers origina... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_partner_churn). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-h-trap-partner-churn",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, h_trap_partner_churn).",
"authors": [],
"... | {
"seed": 250,
"fact_id": "h_trap_partner_churn",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-h-trap-partner-churn |
alabamamikertr/dd-s0250-largest-litigation-claim | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the largest amount claimed (USD) in any pending l... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, largest_litigation_claim). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-largest-litigation-claim",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, largest_litigation_claim).",
"authors": [... | {
"seed": 250,
"fact_id": "largest_litigation_claim",
"tier": 1,
"question_type": "lookup",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-largest-litigation-claim |
alabamamikertr/dd-s0250-net-revenue-retention | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 net revenue retention (FY2025 ARR from cu... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, net_revenue_retention). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-net-revenue-retention",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, net_revenue_retention).",
"authors": [],
... | {
"seed": 250,
"fact_id": "net_revenue_retention",
"tier": 3,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-net-revenue-retention |
alabamamikertr/dd-s0250-pdf-pipeline | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What qualified sales pipeline (USD) did management report... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, pdf_pipeline). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"lookup"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-pdf-pipeline",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, pdf_pipeline).",
"authors": [],
"keywords": [
... | {
"seed": 250,
"fact_id": "pdf_pipeline",
"tier": 2,
"question_type": "lookup",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-pdf-pipeline |
alabamamikertr/dd-s0250-revenue-cagr-23-25 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the FY2023-FY2025 revenue CAGR, as a percentage? ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, revenue_cagr_23_25). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-revenue-cagr-23-25",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, revenue_cagr_23_25).",
"authors": [],
"keyw... | {
"seed": 250,
"fact_id": "revenue_cagr_23_25",
"tier": 2,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-revenue-cagr-23-25 |
alabamamikertr/dd-s0250-top5-concentration | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What share of total FY2025 ARR comes from the five larges... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, top5_concentration). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-top5-concentration",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, top5_concentration).",
"authors": [],
"keyw... | {
"seed": 250,
"fact_id": "top5_concentration",
"tier": 2,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-top5-concentration |
alabamamikertr/dd-s0250-total-headcount | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was total FTE headcount at December 31, 2025?
Answe... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, total_headcount). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup_calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-total-headcount",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, total_headcount).",
"authors": [],
"keywords":... | {
"seed": 250,
"fact_id": "total_headcount",
"tier": 1,
"question_type": "lookup_calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-total-headcount |
alabamamikertr/dd-s0250-trap-cac-payback | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the company's CAC payback period in months?
Answ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, trap_cac_payback). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-trap-cac-payback",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, trap_cac_payback).",
"authors": [],
"keywords... | {
"seed": 250,
"fact_id": "trap_cac_payback",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-trap-cac-payback |
alabamamikertr/dd-s0250-trap-cohort-churn | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the logo churn rate of the 2022 customer cohort ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, trap_cohort_churn). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-trap-cohort-churn",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, trap_cohort_churn).",
"authors": [],
"keywor... | {
"seed": 250,
"fact_id": "trap_cohort_churn",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-trap-cohort-churn |
alabamamikertr/dd-s0250-trap-fy2022-revenue | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the company's FY2022 revenue (USD)?
Answer form... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, trap_fy2022_revenue). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0250-trap-fy2022-revenue",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 250, trap_fy2022_revenue).",
"authors": [],
"ke... | {
"seed": 250,
"fact_id": "trap_fy2022_revenue",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0250-trap-fy2022-revenue |
alabamamikertr/dd-s0251-arr-ex-coc | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is total FY2025 ARR excluding customers whose execut... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, arr_ex_coc). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-arr-ex-coc",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, arr_ex_coc).",
"authors": [],
"keywords": [
"... | {
"seed": 251,
"fact_id": "arr_ex_coc",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-arr-ex-coc |
alabamamikertr/dd-s0251-budget-growth-fy26 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What FY2026 revenue growth does management's budget imply... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, budget_growth_fy26). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-budget-growth-fy26",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, budget_growth_fy26).",
"authors": [],
"keyw... | {
"seed": 251,
"fact_id": "budget_growth_fy26",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-budget-growth-fy26 |
alabamamikertr/dd-s0251-cim-revenue-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does FY2025 revenue stated in the CIM e... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, cim_revenue_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"reconciliation"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-cim-revenue-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, cim_revenue_gap).",
"authors": [],
"keywords":... | {
"seed": 251,
"fact_id": "cim_revenue_gap",
"tier": 2,
"question_type": "reconciliation",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-cim-revenue-gap |
alabamamikertr/dd-s0251-coc-customers | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Which of the company's customers have a right under their... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, coc_customers). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"red_flag"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-coc-customers",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, coc_customers).",
"authors": [],
"keywords": [
... | {
"seed": 251,
"fact_id": "coc_customers",
"tier": 3,
"question_type": "red_flag",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-coc-customers |
alabamamikertr/dd-s0251-fy24-gross-margin | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the company's FY2024 gross margin (gross profit ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, fy24_gross_margin). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup_calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-fy24-gross-margin",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, fy24_gross_margin).",
"authors": [],
"keywor... | {
"seed": 251,
"fact_id": "fy24_gross_margin",
"tier": 1,
"question_type": "lookup_calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-fy24-gross-margin |
alabamamikertr/dd-s0251-h-cim-ebitda-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does the Adjusted EBITDA stated in the ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_cim_ebitda_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"reconciliation_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-cim-ebitda-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_cim_ebitda_gap).",
"authors": [],
"keywords... | {
"seed": 251,
"fact_id": "h_cim_ebitda_gap",
"tier": 4,
"question_type": "reconciliation_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-cim-ebitda-gap |
alabamamikertr/dd-s0251-h-coc-h1-exposure | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Customers whose executed MSA allows termination on a chan... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_coc_h1_exposure). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-coc-h1-exposure",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_coc_h1_exposure).",
"authors": [],
"keywor... | {
"seed": 251,
"fact_id": "h_coc_h1_exposure",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-coc-h1-exposure |
alabamamikertr/dd-s0251-h-diligence-ebitda | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is diligence-adjusted FY2025 EBITDA (USD): EBITDA pe... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_diligence_ebitda). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-diligence-ebitda",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_diligence_ebitda).",
"authors": [],
"keyw... | {
"seed": 251,
"fact_id": "h_diligence_ebitda",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-diligence-ebitda |
alabamamikertr/dd-s0251-h-latest-renewal | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Per the FY2026 renewal schedule, which customer (by name)... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_latest_renewal). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"pdf_table_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-latest-renewal",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_latest_renewal).",
"authors": [],
"keywords... | {
"seed": 251,
"fact_id": "h_latest_renewal",
"tier": 4,
"question_type": "pdf_table_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-latest-renewal |
alabamamikertr/dd-s0251-h-top5-enterprise-precedence | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Applying the data room's source-precedence rules for cust... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_top5_enterprise_precedence). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"calc_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-top5-enterprise-precedence",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_top5_enterprise_precedence).",
"aut... | {
"seed": 251,
"fact_id": "h_top5_enterprise_precedence",
"tier": 4,
"question_type": "calc_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-top5-enterprise-precedence |
alabamamikertr/dd-s0251-h-trap-contract-len-2023 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the average initial contract term, in months, for... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_contract_len_2023). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-trap-contract-len-2023",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_contract_len_2023).",
"authors": [... | {
"seed": 251,
"fact_id": "h_trap_contract_len_2023",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-trap-contract-len-2023 |
alabamamikertr/dd-s0251-h-trap-expansion-h2-2025-renewals | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** How much expansion ARR (USD) came from customers whose co... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_expansion_h2_2025_renewals). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-trap-expansion-h2-2025-renewals",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_expansion_h2_2025_renewals)."... | {
"seed": 251,
"fact_id": "h_trap_expansion_h2_2025_renewals",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-trap-expansion-h2-2025-renewals |
alabamamikertr/dd-s0251-h-trap-grr-fy24-cohort | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 gross revenue retention for customers fir... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_grr_fy24_cohort). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-trap-grr-fy24-cohort",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_grr_fy24_cohort).",
"authors": [],
... | {
"seed": 251,
"fact_id": "h_trap_grr_fy24_cohort",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-trap-grr-fy24-cohort |
alabamamikertr/dd-s0251-h-trap-nrr-ent-fy23-cohort | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 net revenue retention for Enterprise cust... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_nrr_ent_fy23_cohort). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-trap-nrr-ent-fy23-cohort",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_nrr_ent_fy23_cohort).",
"authors... | {
"seed": 251,
"fact_id": "h_trap_nrr_ent_fy23_cohort",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-trap-nrr-ent-fy23-cohort |
alabamamikertr/dd-s0251-h-trap-partner-churn | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** How much FY2025 ARR (USD) was lost from customers origina... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_partner_churn). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-h-trap-partner-churn",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, h_trap_partner_churn).",
"authors": [],
"... | {
"seed": 251,
"fact_id": "h_trap_partner_churn",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-h-trap-partner-churn |
alabamamikertr/dd-s0251-largest-litigation-claim | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the largest amount claimed (USD) in any pending l... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, largest_litigation_claim). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-largest-litigation-claim",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, largest_litigation_claim).",
"authors": [... | {
"seed": 251,
"fact_id": "largest_litigation_claim",
"tier": 1,
"question_type": "lookup",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-largest-litigation-claim |
alabamamikertr/dd-s0251-net-revenue-retention | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 net revenue retention (FY2025 ARR from cu... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, net_revenue_retention). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-net-revenue-retention",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, net_revenue_retention).",
"authors": [],
... | {
"seed": 251,
"fact_id": "net_revenue_retention",
"tier": 3,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-net-revenue-retention |
alabamamikertr/dd-s0251-pdf-pipeline | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What qualified sales pipeline (USD) did management report... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, pdf_pipeline). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"lookup"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-pdf-pipeline",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, pdf_pipeline).",
"authors": [],
"keywords": [
... | {
"seed": 251,
"fact_id": "pdf_pipeline",
"tier": 2,
"question_type": "lookup",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-pdf-pipeline |
alabamamikertr/dd-s0251-revenue-cagr-23-25 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the FY2023-FY2025 revenue CAGR, as a percentage? ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, revenue_cagr_23_25). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-revenue-cagr-23-25",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, revenue_cagr_23_25).",
"authors": [],
"keyw... | {
"seed": 251,
"fact_id": "revenue_cagr_23_25",
"tier": 2,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-revenue-cagr-23-25 |
alabamamikertr/dd-s0251-top5-concentration | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What share of total FY2025 ARR comes from the five larges... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, top5_concentration). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-top5-concentration",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, top5_concentration).",
"authors": [],
"keyw... | {
"seed": 251,
"fact_id": "top5_concentration",
"tier": 2,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-top5-concentration |
alabamamikertr/dd-s0251-total-headcount | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was total FTE headcount at December 31, 2025?
Answe... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, total_headcount). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup_calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-total-headcount",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, total_headcount).",
"authors": [],
"keywords":... | {
"seed": 251,
"fact_id": "total_headcount",
"tier": 1,
"question_type": "lookup_calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-total-headcount |
alabamamikertr/dd-s0251-trap-cac-payback | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the company's CAC payback period in months?
Answ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, trap_cac_payback). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-trap-cac-payback",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, trap_cac_payback).",
"authors": [],
"keywords... | {
"seed": 251,
"fact_id": "trap_cac_payback",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-trap-cac-payback |
alabamamikertr/dd-s0251-trap-cohort-churn | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the logo churn rate of the 2022 customer cohort ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, trap_cohort_churn). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-trap-cohort-churn",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, trap_cohort_churn).",
"authors": [],
"keywor... | {
"seed": 251,
"fact_id": "trap_cohort_churn",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-trap-cohort-churn |
alabamamikertr/dd-s0251-trap-fy2022-revenue | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the company's FY2022 revenue (USD)?
Answer form... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, trap_fy2022_revenue). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0251-trap-fy2022-revenue",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 251, trap_fy2022_revenue).",
"authors": [],
"ke... | {
"seed": 251,
"fact_id": "trap_fy2022_revenue",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0251-trap-fy2022-revenue |
alabamamikertr/dd-s0252-arr-ex-coc | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is total FY2025 ARR excluding customers whose execut... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, arr_ex_coc). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-arr-ex-coc",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, arr_ex_coc).",
"authors": [],
"keywords": [
"... | {
"seed": 252,
"fact_id": "arr_ex_coc",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-arr-ex-coc |
alabamamikertr/dd-s0252-budget-growth-fy26 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What FY2026 revenue growth does management's budget imply... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, budget_growth_fy26). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-budget-growth-fy26",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, budget_growth_fy26).",
"authors": [],
"keyw... | {
"seed": 252,
"fact_id": "budget_growth_fy26",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-budget-growth-fy26 |
alabamamikertr/dd-s0252-cim-revenue-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does FY2025 revenue stated in the CIM e... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, cim_revenue_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"reconciliation"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-cim-revenue-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, cim_revenue_gap).",
"authors": [],
"keywords":... | {
"seed": 252,
"fact_id": "cim_revenue_gap",
"tier": 2,
"question_type": "reconciliation",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-cim-revenue-gap |
alabamamikertr/dd-s0252-coc-customers | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Which of the company's customers have a right under their... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, coc_customers). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"red_flag"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-coc-customers",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, coc_customers).",
"authors": [],
"keywords": [
... | {
"seed": 252,
"fact_id": "coc_customers",
"tier": 3,
"question_type": "red_flag",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-coc-customers |
alabamamikertr/dd-s0252-fy24-gross-margin | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the company's FY2024 gross margin (gross profit ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, fy24_gross_margin). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup_calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-fy24-gross-margin",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, fy24_gross_margin).",
"authors": [],
"keywor... | {
"seed": 252,
"fact_id": "fy24_gross_margin",
"tier": 1,
"question_type": "lookup_calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-fy24-gross-margin |
alabamamikertr/dd-s0252-h-cim-ebitda-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does the Adjusted EBITDA stated in the ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_cim_ebitda_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"reconciliation_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-cim-ebitda-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_cim_ebitda_gap).",
"authors": [],
"keywords... | {
"seed": 252,
"fact_id": "h_cim_ebitda_gap",
"tier": 4,
"question_type": "reconciliation_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-cim-ebitda-gap |
alabamamikertr/dd-s0252-h-coc-h1-exposure | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Customers whose executed MSA allows termination on a chan... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_coc_h1_exposure). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-coc-h1-exposure",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_coc_h1_exposure).",
"authors": [],
"keywor... | {
"seed": 252,
"fact_id": "h_coc_h1_exposure",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-coc-h1-exposure |
alabamamikertr/dd-s0252-h-diligence-ebitda | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is diligence-adjusted FY2025 EBITDA (USD): EBITDA pe... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_diligence_ebitda). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-diligence-ebitda",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_diligence_ebitda).",
"authors": [],
"keyw... | {
"seed": 252,
"fact_id": "h_diligence_ebitda",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-diligence-ebitda |
alabamamikertr/dd-s0252-h-latest-renewal | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Per the FY2026 renewal schedule, which customer (by name)... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_latest_renewal). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"pdf_table_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-latest-renewal",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_latest_renewal).",
"authors": [],
"keywords... | {
"seed": 252,
"fact_id": "h_latest_renewal",
"tier": 4,
"question_type": "pdf_table_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-latest-renewal |
alabamamikertr/dd-s0252-h-top5-enterprise-precedence | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Applying the data room's source-precedence rules for cust... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_top5_enterprise_precedence). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"calc_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-top5-enterprise-precedence",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_top5_enterprise_precedence).",
"aut... | {
"seed": 252,
"fact_id": "h_top5_enterprise_precedence",
"tier": 4,
"question_type": "calc_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-top5-enterprise-precedence |
alabamamikertr/dd-s0252-h-trap-contract-len-2023 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the average initial contract term, in months, for... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_contract_len_2023). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-trap-contract-len-2023",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_contract_len_2023).",
"authors": [... | {
"seed": 252,
"fact_id": "h_trap_contract_len_2023",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-trap-contract-len-2023 |
alabamamikertr/dd-s0252-h-trap-expansion-h2-2025-renewals | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** How much expansion ARR (USD) came from customers whose co... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_expansion_h2_2025_renewals). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-trap-expansion-h2-2025-renewals",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_expansion_h2_2025_renewals)."... | {
"seed": 252,
"fact_id": "h_trap_expansion_h2_2025_renewals",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-trap-expansion-h2-2025-renewals |
alabamamikertr/dd-s0252-h-trap-grr-fy24-cohort | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 gross revenue retention for customers fir... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_grr_fy24_cohort). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-trap-grr-fy24-cohort",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_grr_fy24_cohort).",
"authors": [],
... | {
"seed": 252,
"fact_id": "h_trap_grr_fy24_cohort",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-trap-grr-fy24-cohort |
alabamamikertr/dd-s0252-h-trap-nrr-ent-fy23-cohort | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 net revenue retention for Enterprise cust... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_nrr_ent_fy23_cohort). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-trap-nrr-ent-fy23-cohort",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_nrr_ent_fy23_cohort).",
"authors... | {
"seed": 252,
"fact_id": "h_trap_nrr_ent_fy23_cohort",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-trap-nrr-ent-fy23-cohort |
alabamamikertr/dd-s0252-h-trap-partner-churn | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** How much FY2025 ARR (USD) was lost from customers origina... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_partner_churn). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"trap_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-h-trap-partner-churn",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, h_trap_partner_churn).",
"authors": [],
"... | {
"seed": 252,
"fact_id": "h_trap_partner_churn",
"tier": 4,
"question_type": "trap_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-h-trap-partner-churn |
alabamamikertr/dd-s0252-largest-litigation-claim | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the largest amount claimed (USD) in any pending l... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, largest_litigation_claim). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-largest-litigation-claim",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, largest_litigation_claim).",
"authors": [... | {
"seed": 252,
"fact_id": "largest_litigation_claim",
"tier": 1,
"question_type": "lookup",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-largest-litigation-claim |
alabamamikertr/dd-s0252-net-revenue-retention | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was FY2025 net revenue retention (FY2025 ARR from cu... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, net_revenue_retention). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-net-revenue-retention",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, net_revenue_retention).",
"authors": [],
... | {
"seed": 252,
"fact_id": "net_revenue_retention",
"tier": 3,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-net-revenue-retention |
alabamamikertr/dd-s0252-pdf-pipeline | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What qualified sales pipeline (USD) did management report... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, pdf_pipeline). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"lookup"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-pdf-pipeline",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, pdf_pipeline).",
"authors": [],
"keywords": [
... | {
"seed": 252,
"fact_id": "pdf_pipeline",
"tier": 2,
"question_type": "lookup",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-pdf-pipeline |
alabamamikertr/dd-s0252-revenue-cagr-23-25 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the FY2023-FY2025 revenue CAGR, as a percentage? ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, revenue_cagr_23_25). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-revenue-cagr-23-25",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, revenue_cagr_23_25).",
"authors": [],
"keyw... | {
"seed": 252,
"fact_id": "revenue_cagr_23_25",
"tier": 2,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-revenue-cagr-23-25 |
alabamamikertr/dd-s0252-top5-concentration | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What share of total FY2025 ARR comes from the five larges... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, top5_concentration). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-top5-concentration",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, top5_concentration).",
"authors": [],
"keyw... | {
"seed": 252,
"fact_id": "top5_concentration",
"tier": 2,
"question_type": "calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-top5-concentration |
alabamamikertr/dd-s0252-total-headcount | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was total FTE headcount at December 31, 2025?
Answe... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, total_headcount). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup_calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-total-headcount",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, total_headcount).",
"authors": [],
"keywords":... | {
"seed": 252,
"fact_id": "total_headcount",
"tier": 1,
"question_type": "lookup_calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-total-headcount |
alabamamikertr/dd-s0252-trap-cac-payback | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is the company's CAC payback period in months?
Answ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, trap_cac_payback). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-trap-cac-payback",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, trap_cac_payback).",
"authors": [],
"keywords... | {
"seed": 252,
"fact_id": "trap_cac_payback",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-trap-cac-payback |
alabamamikertr/dd-s0252-trap-cohort-churn | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the logo churn rate of the 2022 customer cohort ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, trap_cohort_churn). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-trap-cohort-churn",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, trap_cohort_churn).",
"authors": [],
"keywor... | {
"seed": 252,
"fact_id": "trap_cohort_churn",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-trap-cohort-churn |
alabamamikertr/dd-s0252-trap-fy2022-revenue | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the company's FY2022 revenue (USD)?
Answer form... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, trap_fy2022_revenue). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"trap"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0252-trap-fy2022-revenue",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 252, trap_fy2022_revenue).",
"authors": [],
"ke... | {
"seed": 252,
"fact_id": "trap_fy2022_revenue",
"tier": 2,
"question_type": "trap",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0252-trap-fy2022-revenue |
alabamamikertr/dd-s0253-arr-ex-coc | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is total FY2025 ARR excluding customers whose execut... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, arr_ex_coc). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-arr-ex-coc",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, arr_ex_coc).",
"authors": [],
"keywords": [
"... | {
"seed": 253,
"fact_id": "arr_ex_coc",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-arr-ex-coc |
alabamamikertr/dd-s0253-budget-growth-fy26 | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What FY2026 revenue growth does management's budget imply... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, budget_growth_fy26). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"multi_hop"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-budget-growth-fy26",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, budget_growth_fy26).",
"authors": [],
"keyw... | {
"seed": 253,
"fact_id": "budget_growth_fy26",
"tier": 3,
"question_type": "multi_hop",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-budget-growth-fy26 |
alabamamikertr/dd-s0253-cim-revenue-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does FY2025 revenue stated in the CIM e... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, cim_revenue_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-2",
"reconciliation"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-cim-revenue-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, cim_revenue_gap).",
"authors": [],
"keywords":... | {
"seed": 253,
"fact_id": "cim_revenue_gap",
"tier": 2,
"question_type": "reconciliation",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-cim-revenue-gap |
alabamamikertr/dd-s0253-coc-customers | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Which of the company's customers have a right under their... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, coc_customers). | [
"due-diligence",
"finance",
"beam-dd",
"tier-3",
"red_flag"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-coc-customers",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, coc_customers).",
"authors": [],
"keywords": [
... | {
"seed": 253,
"fact_id": "coc_customers",
"tier": 3,
"question_type": "red_flag",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-coc-customers |
alabamamikertr/dd-s0253-fy24-gross-margin | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What was the company's FY2024 gross margin (gross profit ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, fy24_gross_margin). | [
"due-diligence",
"finance",
"beam-dd",
"tier-1",
"lookup_calc"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-fy24-gross-margin",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, fy24_gross_margin).",
"authors": [],
"keywor... | {
"seed": 253,
"fact_id": "fy24_gross_margin",
"tier": 1,
"question_type": "lookup_calc",
"suite": "v0.2",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-fy24-gross-margin |
alabamamikertr/dd-s0253-h-cim-ebitda-gap | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** By how much (USD) does the Adjusted EBITDA stated in the ... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_cim_ebitda_gap). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"reconciliation_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-h-cim-ebitda-gap",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_cim_ebitda_gap).",
"authors": [],
"keywords... | {
"seed": 253,
"fact_id": "h_cim_ebitda_gap",
"tier": 4,
"question_type": "reconciliation_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-h-cim-ebitda-gap |
alabamamikertr/dd-s0253-h-coc-h1-exposure | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Customers whose executed MSA allows termination on a chan... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_coc_h1_exposure). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-h-coc-h1-exposure",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_coc_h1_exposure).",
"authors": [],
"keywor... | {
"seed": 253,
"fact_id": "h_coc_h1_exposure",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-h-coc-h1-exposure |
alabamamikertr/dd-s0253-h-diligence-ebitda | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** What is diligence-adjusted FY2025 EBITDA (USD): EBITDA pe... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_diligence_ebitda). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"multi_hop_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-h-diligence-ebitda",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_diligence_ebitda).",
"authors": [],
"keyw... | {
"seed": 253,
"fact_id": "h_diligence_ebitda",
"tier": 4,
"question_type": "multi_hop_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-h-diligence-ebitda |
alabamamikertr/dd-s0253-h-latest-renewal | # Commercial due diligence question
You are an analyst on a buy-side due diligence team. The target company's virtual data room is at `/app/dataroom`
(markdown, CSV and PDF files). Use the shell (ls, cat, grep, python3, pdftotext) to investigate.
**Question:** Per the FY2026 renewal schedule, which customer (by name)... | Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_latest_renewal). | [
"due-diligence",
"finance",
"beam-dd",
"tier-4",
"pdf_table_hard"
] | 1.4 | {
"schema_version": "1.4",
"artifacts": [
"/app/answer.json"
],
"task": {
"name": "alabamamikertr/dd-s0253-h-latest-renewal",
"version": "0.1.0",
"description": "Beam-DD: commercial due-diligence question over a synthetic data room (seed 253, h_latest_renewal).",
"authors": [],
"keywords... | {
"seed": 253,
"fact_id": "h_latest_renewal",
"tier": 4,
"question_type": "pdf_table_hard",
"suite": "hard",
"room": "v0.3-hard"
} | [
"environment/Dockerfile",
"environment/dataroom.tar.gz",
"instruction.md",
"solution/solve.sh",
"task.toml",
"tests/Dockerfile",
"tests/dataroom.tar.gz",
"tests/fact.json",
"tests/grade.py",
"tests/test.sh"
] | frozen/test_hard/dd-s0253-h-latest-renewal |
End of preview. Expand in Data Studio
Dataset card: Beam-DD
Summary
This is a Harbor environment of 11,400 single-question due-diligence tasks over synthetic company data rooms. The agent has a shell and must write /app/answer.json (value, status, citations). Scoring is deterministic: 0.7 for correctness and 0.3 for citations, with abstention required on trap questions. The baseline model is Reflection's Beam-1.0-501B-A23B. Setup and run commands are in README.md.
- Curator: Michael Stricklen (HF
alabamamikertr) - Repo:
alabamamikertr/beam-dd-dataroom; demo Space:alabamamikertr/beam-dd-demo - Generator: v0.4 (52a7d31). Packaging: v0.4.1 (43b9628).
Layout of this repo
frozen/test_hard,frozen/test_normal: Harbor task directories.archives/train_hard.tar,archives/train_normal.tar: the train splits as plain tar (to stay under the Hub file-count limit). Extract withtar -C frozen -xf archives/train_hard.tar, then checkgen/check_tree_hash.sh frozen/train_hard 31224ee8067ae6a4.gen/,base/: generator, grader and agent base image.PROJECT_README.md: full README with Harbor commands and baseline.
Structure
Each task is a Harbor task directory:
instruction.md: the question, answer format and citation rulestask.toml: metadataseed,fact_id,tier,question_type,suite,room;network_mode = "no-network";artifacts = ["/app/answer.json"]environment/: Dockerfile anddataroom.tar.gz, unpacked to/app/dataroomtests/:fact.json(answer key and evidence locations),grade.py, and the pristine room. These go only to the verifier image.solution/solve.sh: the oracle
fact.json schema: {id, type, tier, question, answer: {kind: number|set|bool|string|none, value, unit, tol_rel, tol_abs}, evidence: [{file, locator: {line|row|page}, snippet, group?}]}.
Splits
| split | seeds | normal tasks | hard tasks | tree hashes (normal / hard) |
|---|---|---|---|---|
| train | 0-249 | 3,500 | 6,000 | e0da6a6b191b15a2 / 31224ee8067ae6a4 |
| test | 250-299 | 700 | 1,200 | cc4f39e78960544c / 914742567f50fcf4 |
Per seed there are 14 normal tasks and 24 hard tasks. The hard tasks are the 14 normal questions recomputed under the room's precedence rules, plus 10 tier-4 questions, 5 of which are traps. A dev set (seeds 1000-1003, 56 + 96 tasks) was used for iteration and is not part of the release. The test split is held out by seed and shares templates with train.
Composition per seed (hard room)
| tier | count | types |
|---|---|---|
| 1 | 3 | lookup |
| 2 | 7 | calc, lookup_calc, reconciliation |
| 3 | 4 | multi_hop, red_flag, trap |
| 4 | 10 | calc_hard, multi_hop_hard ×2, reconciliation_hard, pdf_table_hard, trap_hard ×5 |
Traps are 8 of the 24 hard questions per seed (3 from v0.2, 5 tier-4 traps).
Generation
The pipeline is pure Python with fixed seeds (Faker and reportlab with invariant=1), and each split rebuilds byte-for-byte. Facts are sampled first and the documents are rendered from them. Inconsistencies are injected on purpose:
- CIM inflation
- superseded management drafts
- unexecuted contract drafts
- a withdrawn process-letter version
- lookalike-digit OCR errors
- order-form ARR overrides
Names are synthetic. Any resemblance to real entities is coincidental.
Harbor Hub
Both test splits are public on the Harbor Hub. No login is needed to pull them:
alabamamike/beam-dd-test-hard: 1,200 tasks, dataset hash370e91866f47alabamamike/beam-dd-test-normal: 700 tasks, dataset hash350d29662719
harbor dataset download alabamamike/beam-dd-test-hard -o tasks/ # fetch the tasks locally
harbor run -d alabamamike/beam-dd-test-hard -a oracle -n 8 -o jobs # run straight from the Hub
The Hub copies have the same content as frozen/test_*, except for one change: each task's [task].name is prefixed for the Hub namespace, alabamamike/beam-dd-{hard,normal}-<task>. This keeps hard and normal tasks from colliding, because the two splits use the same per-seed task names. That one field means the Hub copies do not match the frozen/ tree hashes. Use the HF dataset for hash-verified files.
Intended use
- Evaluating tool-using agents on document-heavy, rule-following financial QA with abstention.
- RL or rejection-sampling training on the train split. Rewards are dense in [0, 1], and Harbor can collect token logprobs.
The dataset should not be used as evidence of real-world diligence competence, or for any decision about real companies.
Baseline (Beam-1.0-501B-A23B, OpenRouter Nebius fp8, terminus-2, max_turns 30)
| split | effort | n | reward | T4 | traps | non-trap |
|---|---|---|---|---|---|---|
| test-hard | low | 1200 | 0.777 | 0.630 | 0.585 | 0.873 |
| test-normal | low | 700 | 0.981 | — | 0.973 | 0.982 |
| test-hard | medium | 1200 | not run | — | — | — |
| dev-hard k=3 | low | 96 × 3 | 0.808 ± 0.004 | 0.688 | 0.625 | 0.899 |
1 trial per task on test, 0 errored trials. Reports: reports/test_hard.txt, reports/test_normal.txt, reports/dev_hard_k3.txt; failure analysis reports/failure_analysis.md. Oracle 1.0 / nop 0.0 on a 48-task test-hard slice (seeds 250-251). Verify a split with gen/check_tree_hash.sh frozen/test_hard 914742567f50fcf4.
Limitations and biases
- The rooms are synthetic and template-based: one sector framing, USD and US GAAP, English, and cleaner than real VDRs.
- There is a fixed set of trap templates, so models can overfit to them during training.
- The citation check is positional (±3 lines/rows, or the exact PDF page) and not semantic.
- Tolerances are tight, so some reasonable rounding conventions score 0.
- Single-question tasks do not test long-horizon memo writing.
- Agent network isolation depends on the Harbor runtime.
License
MIT, covering the code and the generated tasks. See LICENSE.
- Downloads last month
- 251