The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
id: string
dataset: struct<meta: struct<generator: string, generatorVersion: string, schemaVersion: string, industry: st (... 4447 chars omitted)
child 0, meta: struct<generator: string, generatorVersion: string, schemaVersion: string, industry: string, seed: i (... 512 chars omitted)
child 0, generator: string
child 1, generatorVersion: string
child 2, schemaVersion: string
child 3, industry: string
child 4, seed: int64
child 5, startDate: timestamp[s]
child 6, asOf: timestamp[s]
child 7, synthetic: bool
child 8, calibration: struct<cbp: struct<year: int64, naics: string, naicsVintage: int64, employees: int64, establishments (... 330 chars omitted)
child 0, cbp: struct<year: int64, naics: string, naicsVintage: int64, employees: int64, establishments: int64, emp (... 31 chars omitted)
child 0, year: int64
child 1, naics: string
child 2, naicsVintage: int64
child 3, employees: int64
child 4, establishments: int64
child 5, employeesPerEstablishment: double
child 1, susb: struct<year: int64, firms: int64, firmsUnder20: int64, smallFirmShare: double>
child 0, year: int64
child 1, firms: int64
child 2, firmsUnder20: int64
child 3, smallFirmShare: double
child 2, oews: struct<year: int64, naics: string, occupation: string, medianHourlyCents: int64>
...
ted)
child 0, item: struct<id: string, sourceType: string, sourceId: string, date: timestamp[s], lines: list<item: struc (... 59 chars omitted)
child 0, id: string
child 1, sourceType: string
child 2, sourceId: string
child 3, date: timestamp[s]
child 4, lines: list<item: struct<account: string, debitCents: int64, creditCents: int64>>
child 0, item: struct<account: string, debitCents: int64, creditCents: int64>
child 0, account: string
child 1, debitCents: int64
child 2, creditCents: int64
inventory: list<item: null>
child 0, item: null
invariants: struct<violations: list<item: struct<rule: string, recordIds: list<item: string>, detail: string>>, (... 130 chars omitted)
child 0, violations: list<item: struct<rule: string, recordIds: list<item: string>, detail: string>>
child 0, item: struct<rule: string, recordIds: list<item: string>, detail: string>
child 0, rule: string
child 1, recordIds: list<item: string>
child 0, item: string
child 2, detail: string
child 1, notChecked: list<item: string>
child 0, item: string
child 2, ledgerViolations: list<item: null>
child 0, item: null
child 3, inventoryViolations: list<item: null>
child 0, item: null
child 4, inventoryScope: string
repository: string
zenodo_doi: string
pull_request: int64
generated_from: string
commit: string
npm_publication: string
hugging_face: string
to
{'repository': Value('string'), 'commit': Value('string'), 'pull_request': Value('int64'), 'generated_from': Value('string'), 'npm_publication': Value('string'), 'hugging_face': Value('string'), 'zenodo_doi': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2951, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2461, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2486, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 547, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 430, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
id: string
dataset: struct<meta: struct<generator: string, generatorVersion: string, schemaVersion: string, industry: st (... 4447 chars omitted)
child 0, meta: struct<generator: string, generatorVersion: string, schemaVersion: string, industry: string, seed: i (... 512 chars omitted)
child 0, generator: string
child 1, generatorVersion: string
child 2, schemaVersion: string
child 3, industry: string
child 4, seed: int64
child 5, startDate: timestamp[s]
child 6, asOf: timestamp[s]
child 7, synthetic: bool
child 8, calibration: struct<cbp: struct<year: int64, naics: string, naicsVintage: int64, employees: int64, establishments (... 330 chars omitted)
child 0, cbp: struct<year: int64, naics: string, naicsVintage: int64, employees: int64, establishments: int64, emp (... 31 chars omitted)
child 0, year: int64
child 1, naics: string
child 2, naicsVintage: int64
child 3, employees: int64
child 4, establishments: int64
child 5, employeesPerEstablishment: double
child 1, susb: struct<year: int64, firms: int64, firmsUnder20: int64, smallFirmShare: double>
child 0, year: int64
child 1, firms: int64
child 2, firmsUnder20: int64
child 3, smallFirmShare: double
child 2, oews: struct<year: int64, naics: string, occupation: string, medianHourlyCents: int64>
...
ted)
child 0, item: struct<id: string, sourceType: string, sourceId: string, date: timestamp[s], lines: list<item: struc (... 59 chars omitted)
child 0, id: string
child 1, sourceType: string
child 2, sourceId: string
child 3, date: timestamp[s]
child 4, lines: list<item: struct<account: string, debitCents: int64, creditCents: int64>>
child 0, item: struct<account: string, debitCents: int64, creditCents: int64>
child 0, account: string
child 1, debitCents: int64
child 2, creditCents: int64
inventory: list<item: null>
child 0, item: null
invariants: struct<violations: list<item: struct<rule: string, recordIds: list<item: string>, detail: string>>, (... 130 chars omitted)
child 0, violations: list<item: struct<rule: string, recordIds: list<item: string>, detail: string>>
child 0, item: struct<rule: string, recordIds: list<item: string>, detail: string>
child 0, rule: string
child 1, recordIds: list<item: string>
child 0, item: string
child 2, detail: string
child 1, notChecked: list<item: string>
child 0, item: string
child 2, ledgerViolations: list<item: null>
child 0, item: null
child 3, inventoryViolations: list<item: null>
child 0, item: null
child 4, inventoryScope: string
repository: string
zenodo_doi: string
pull_request: int64
generated_from: string
commit: string
npm_publication: string
hugging_face: string
to
{'repository': Value('string'), 'commit': Value('string'), 'pull_request': Value('int64'), 'generated_from': Value('string'), 'npm_publication': Value('string'), 'hugging_face': Value('string'), 'zenodo_doi': Value('string')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
TeamShift SMB Sandbox Three-Year Company Library
100 deterministic fictional companies spanning home services, dental clinics, marketing agencies and independent trailer dealers. Each spans 2023-10-01 through the simulated as-of date 2026-09-30, plus explicitly scheduled future appointments. This is synthetic test data, not sampled customer data or a measured business-outcome benchmark.
Build the existing workspace, then run node scripts/export-library.mjs artifacts/library-v1. Each gzipped JSON file contains the original linked business dataset, a mock double-entry ledger, modeled trailer stock movements and an invariant report. manifest.json supplies generation options and SHA-256 checksums. fictional-names.json records generated company names; callers can use checkFictionalNames with an explicit registry to screen near matches. Screening is not trademark clearance.
Email domains use reserved .example and telephone numbers use fictional 555-01xx ranges. Human-readable names are invented and may coincidentally resemble real entities. Do not contact them, use them for eligibility decisions, or treat them as customer facts.
Calibration and limitations
Metadata records Census CBP 2023 employment and establishment counts, SUSB 2022 enterprise-size counts and BLS OEWS May 2023 occupational wages. Medium active staffing is rounded from CBP employment divided by establishments. SUSB shares and OEWS wages are contextual references, not a fitted distribution or generated payroll. Other sizes, role mixes, sales prices, appointment volumes, conversion rates and event timings are illustrative synthetic settings.
Sources: CBP national data, SUSB detailed enterprise sizes, SUSB methodology, and BLS tables linked per profile in metadata. CBP/SUSB use 2017 NAICS. Marketing wages use broader 541800. Utility trailer dealers use 2017 441228 and broader OEWS441200 as proxies; 2022 utility-trailer classification is broader 441227. These employer aggregates include larger firms and exclude self-employed workers; they do not establish SMB-only realism.
Defects and accounting
Deliberate, labeled anomalies include invoice/job mismatches and conflicting statuses. Invariant reports retain those defects. The illustrative ledger books non-draft/non-void invoices into receivables/sales and received payments into cash/receivables or unapplied cash. It is not tax advice or a production posting system. Trailer opening stock is an explicit synthetic assumption sufficient for generated completed sales, not measured inventory. Other industries have no modeled stock, and inventory remains marked untested there.
Reproduction and publication
fake-business --list-companies lists the stable IDs; fake-business --company company-001 regenerates an entry. The library configurations opt into 36 months and calibration without changing generate() defaults. Do not combine a library ID with generation overrides.
License: the repository's Apache-2.0 license. The reviewed 100-company archive is published on Zenodo as library-v1; its SHA-256 is 038a982f87ed115ad6e271229c13240cf95e2c576d1280649b4dbecdb23069ad. Use the dataset citation for that deposit. The repository-root citation remains the software's 0.1.0 citation. The same files are mirrored on Hugging Face.
- Downloads last month
- 42