Dataset Viewer
Auto-converted to Parquet Duplicate
query
stringlengths
9
691
route
listlengths
2
7
context
nullclasses
22 values
Review internal workflow documentation and current workflow frameworks before generating a workflow engine.
[ [ "RAG" ], [ "WEB_SEARCH" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Analyze an uploaded earnings report using internal analyst notes and recent news
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: We maintain internal examples for third-party integrations. Use them to implement a new CRM API integration in the uploaded backend.
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: The uploaded Node.js app currently stores payments manually. Integrate PayPal checkout and webhook reconciliation.
[ [ "FILE_CONTEXT" ], [ "WEB_SEARCH" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Research how successful AI products define evaluation metrics in their PRDs.
[ [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Research recent trends in AI startups and recommend promising opportunities.
[ [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Investigate the global data center industry. Research power consumption, AI-driven demand, geographic trends, construction constraints, economics, and future growth projections.
[ [ "WEB_SEARCH" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Analyze the uploaded AI system design using our internal AI guidelines and current industry approaches.
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: The uploaded codebase implements a vector search platform. Internal documents describe earlier search systems and known limitations. The vector database ecosystem has changed rapidly. Analyze the uploaded implementation, compare it against internal lessons and current vector search approaches, and generate re...
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate code for metadata-aware retrieval.
[ [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Using our internal pricing documentation, calculate the cost for 250,000 API calls.
[ [ "RAG" ], [ "TOOL_CALL" ], [ "ANSWER" ], [ "EOF" ] ]
null
What's a good way to learn a new skill quickly?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
How often should I water a houseplant?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare two uploaded code review reports and identify recurring engineering issues.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
The backlog appears to grow faster than I can complete tasks.
[ [ "ANSWER" ], [ "EOF" ] ]
null
Review the assumptions in my uploaded DCF model
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
I've spent months improving the architecture of my application. Everything is cleaner and more flexible than before. The problem is that users don't seem to care about architecture. They only notice visible features. Am I focusing on the wrong things?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Search our internal AI documentation and compare it with recent research
[ [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: I need to port a Python FastAPI endpoint to Go. Convert the endpoint to Go using Gin and equivalent request/response structs.
[ [ "DIRECT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate code for a document chunking evaluation framework.
[ [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare the uploaded contract against recent regulatory requirements in this industry.
[ [ "FILE_CONTEXT" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: We uploaded a cloud migration plan, maintain internal infrastructure standards, and want to understand whether recent cloud provider changes affect the strategy. Compare all available information and provide updated recommendations.
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Use our internal software lifecycle standards to plan a new product.
[ [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: The uploaded backend uses raw HTTP calls everywhere. Introduce a reusable API gateway client layer with rate limiting and structured logging.
[ [ "FILE_CONTEXT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Determine whether a filing deadline has passed
[ [ "TOOL_CALL" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate a software implementation roadmap using our internal development process documentation.
[ [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: The attached pull request introduces websocket communication. Review the implementation for protocol design, reliability, scalability, and error handling.
[ [ "FILE_CONTEXT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
I recently received feedback that my presentations contain too much detail. My instinct is to explain everything thoroughly because I don't want important context to be lost. On the other hand, people seem to disengage before I get to the main point. How can I make my presentations better?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: Several teams have investigated observability and monitoring improvements. We want a consolidated overview. Search internal records and summarize recommendations, implementations, and measured outcomes.
[ [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
Take the roadmap from our last conversation and adjust it for a smaller team.
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Write a Flutter login screen
[ [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: The repository has become difficult to navigate. Refactor the folder structure and module organization.
[ [ "FILE_CONTEXT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Analyze uploaded production logs using internal incident reports and recent vendor advisories
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Summarize the financial statements in my uploaded PDF
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate an AI prompt template manager from these requirements: create templates, version them, assign variables, test outputs, and publish stable versions.
[ [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Analyze my uploaded pricing strategy and recommend improvements.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Analyze the uploaded requirements and build an implementation plan using current SaaS best practices.
[ [ "FILE_CONTEXT" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
What's a good movie to watch tonight?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review my uploaded contractor agreement and identify confidentiality obligations.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
I've uploaded user interview transcripts. There are dozens of conversations to review. Can you summarize common themes?
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Analyze the uploaded spreadsheet and identify the customers generating the most revenue.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Find documentation.
[ [ "CLARIFY" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: An uploaded application is failing after a framework upgrade. We suspect breaking changes in the framework itself. Compare the uploaded implementation against recent release notes and documentation.
[ [ "FILE_CONTEXT" ], [ "WEB_SEARCH" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review it against both our standards and industry best practices.
[ [ "CLARIFY" ], [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: We uploaded a market analysis report and have years of internal strategy documents. Recent industry developments may change the conclusions. Review all sources and explain how the current situation differs from previous assumptions.
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Search internal documents for key management procedures
[ [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
Search our internal startup archive for examples of failed launches.
[ [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
What are the key metrics discussed in this file?
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review uploaded prompts and generate an optimized prompt library.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Design an architecture for an AI-powered customer support platform.
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Summarize my uploaded market research and compare the findings with current industry trends.
[ [ "FILE_CONTEXT" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review our compliance posture.
[ [ "CLARIFY" ], [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Use the same market assumptions as before.
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Create a technical summary of the uploaded API specification.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare two uploaded software design documents and identify architectural differences.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: We uploaded a database schema that has evolved organically over several years. Internal architecture documentation contains recommended modeling patterns, and we want to incorporate modern database design practices. Analyze the uploaded schema, compare it against internal standards and current recommendations...
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
What hardware should I buy for local 70B inference under $5000?
[ [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: A research proposal includes a population growth model and needs numerical validation. Calculate expected population size after a specified number of years using the provided assumptions.
[ [ "TOOL_CALL" ], [ "ANSWER" ], [ "EOF" ] ]
null
Read the uploaded deployment guide and extract deployment steps and prerequisites.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Calculate month-over-month growth from the uploaded spreadsheet.
[ [ "FILE_CONTEXT" ], [ "TOOL_CALL" ], [ "ANSWER" ], [ "EOF" ] ]
null
What are the latest developments in ESG investing?
[ [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Calculate the weighted average of grades 85, 90, and 78 with weights 40%, 35%, and 25%.
[ [ "TOOL_CALL" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review the uploaded product specification and identify missing edge cases.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate a websocket protocol implementation from these requirements: each message has a type, request ID, sender, payload, timestamp, and optional error.
[ [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Continue from the notes we created previously.
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review the uploaded research protocol and suggest improvements based on current standards
[ [ "FILE_CONTEXT" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Calculate the compound annual growth rate from $100,000 to $450,000 over 12 years.
[ [ "TOOL_CALL" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare the uploaded pricing model with current market pricing.
[ [ "FILE_CONTEXT" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Convert 3.2 million seconds into days.
[ [ "TOOL_CALL" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: We uploaded a software architecture document for a new SaaS platform. Internal engineering guidance defines preferred patterns and previous platform implementations. We also want to align with modern cloud-native architectures. Review the uploaded architecture, compare it with internal knowledge and current i...
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
The company I'm working for has gone through several organizational changes. Responsibilities shifted, teams changed, and some long-term plans were adjusted. Nothing is dramatically wrong, but I find myself thinking more about future opportunities. Should I look for a new job?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: We have two uploaded versions of the same product specification created six months apart. Management wants to understand what changed. Compare the uploaded files and summarize additions, removals, modifications, and their potential business impact.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare AI agent platforms by capabilities, pricing, and developer adoption.
[ [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare local GGUF models stored in our internal catalog by inference speed.
[ [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare my uploaded resume with a job description and identify gaps.
[ [ "FILE_CONTEXT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Find internal documentation
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Search internal deployment standards for local AI infrastructure
[ [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
How do I become more confident at work?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Create a Python script to model population growth
[ [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: This repository contains over 300,000 lines of code. Identify architectural hotspots that deserve deeper manual review.
[ [ "FILE_CONTEXT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: This React hook manages websocket connections and retry logic. Convert it to a Flutter GetX controller with equivalent behavior.
[ [ "DIRECT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Research recent startup funding trends in the AI industry.
[ [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate SQL migrations from the uploaded schema document
[ [ "FILE_CONTEXT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate a roadmap for launching a local AI assistant from MVP to enterprise scale.
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Debug the uploaded React project
[ [ "FILE_CONTEXT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review my uploaded requirements, compare them with internal standards, and evaluate them against modern product management practices.
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Analyze the uploaded policy document against internal compliance requirements
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate a roadmap for a startup building AI-powered developer tools.
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Compare our infrastructure against internal standards and industry best practices.
[ [ "CLARIFY" ], [ "FILE_CONTEXT" ], [ "RAG" ], [ "WEB_SEARCH" ], [ "ANSWER" ], [ "EOF" ] ]
null
Can you finish the technical specification you started earlier?
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Convert the uploaded inventory spreadsheet into normalized SQL tables.
[ [ "FILE_CONTEXT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Create a market positioning matrix for local LLM software vendors.
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
What was that list of priorities we made?
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Context: This Go function validates API payloads using structs and custom errors. Convert it to Rust with serde validation and equivalent error handling.
[ [ "DIRECT" ], [ "CODE" ], [ "ANSWER" ], [ "EOF" ] ]
null
Review the uploaded design document using our engineering handbook
[ [ "FILE_CONTEXT" ], [ "RAG" ], [ "ANSWER" ], [ "EOF" ] ]
null
How can I make friends as an adult?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
Estimate the timeline.
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Generate a technical specification.
[ [ "CLARIFY" ], [ "ANSWER" ], [ "EOF" ] ]
null
Create a phased implementation plan for a knowledge management system.
[ [ "ANSWER" ], [ "EOF" ] ]
null
What's a quick lunch I can make in 10 minutes?
[ [ "DIRECT" ], [ "ANSWER" ], [ "EOF" ] ]
null
End of preview. Expand in Data Studio

Wisedome Routing Dataset

Dataset Description

The Wisedome Routing Dataset is a workflow planning and routing dataset designed for training AI assistants, agent systems, RAG pipelines, and workflow orchestrators.

Each example contains a user query and the expected sequence of actions required to fulfill the request. Unlike traditional intent-classification datasets that predict a single label, this dataset focuses on multi-step workflow generation.

The dataset can be used to train models that predict the next action in a workflow or generate complete execution routes.


Dataset Structure

Each example contains:

{
  "query": "Compare the uploaded report with recent market trends",
  "route": [
    ["FILE_CONTEXT"],
    ["WEB_SEARCH"],
    ["ANSWER"],
    ["EOF"]
  ]
}

Fields

Field Description
query User request
route Ordered workflow actions required to satisfy the request

Routes are represented as sequential workflow steps.


Supported Actions

Action Description
DIRECT Answer directly from model knowledge
WEB_SEARCH Retrieve information from the internet
FILE_CONTEXT Retrieve information from uploaded files
RAG Retrieve information from indexed local knowledge
TOOL_CALL Execute a tool, calculation, or external function
CODE Generate code
CLARIFY Request clarification from the user
ANSWER Generate the final response
EOF End workflow

Example Routes

Direct Question

{
  "query": "What is SQLite?",
  "route": [
    ["DIRECT"],
    ["ANSWER"],
    ["EOF"]
  ]
}

Web Research

{
  "query": "What changed in the latest Flutter release?",
  "route": [
    ["WEB_SEARCH"],
    ["ANSWER"],
    ["EOF"]
  ]
}

File Analysis

{
  "query": "Summarize the uploaded PDF",
  "route": [
    ["FILE_CONTEXT"],
    ["ANSWER"],
    ["EOF"]
  ]
}

Combined Retrieval

{
  "query": "Compare the uploaded report with current market news",
  "route": [
    ["FILE_CONTEXT"],
    ["WEB_SEARCH"],
    ["ANSWER"],
    ["EOF"]
  ]
}

Code Generation

{
  "query": "Write a Flutter widget for stacked avatars",
  "route": [
    ["CODE"],
    ["ANSWER"],
    ["EOF"]
  ]
}

Intended Use

The dataset is intended for:

  • Workflow routing
  • Agent planning
  • Retrieval-augmented generation (RAG)
  • Tool-using assistants
  • AI orchestration systems
  • Local AI assistants
  • Multi-step reasoning pipelines

Dataset Statistics

Total Examples

  • Total examples: INSERT_COUNT
  • Training examples: INSERT_TRAIN_COUNT
  • Test examples: INSERT_TEST_COUNT

Action Distribution

Action Count
ANSWER 5188
EOF 5188
FILE_CONTEXT 2428
WEB_SEARCH 1359
CODE 1202
RAG 1106
DIRECT 685
CLARIFY 661
TOOL_CALL 392

Route Statistics

  • Average route length: 3.51
  • Maximum route length: 7
  • Minimum route length: 2

Training Use

The dataset can be used in two ways:

Route Prediction

Train models to generate complete workflow routes from user requests.

Next-Step Prediction

Convert routes into training examples where the model predicts the next action given:

  • User query
  • Previous workflow actions
  • Optional context state

Example:

Input:

Query: Compare the uploaded report with current news
Previous actions: FILE_CONTEXT

Target:

WEB_SEARCH

Limitations

The dataset reflects workflow definitions used by the Wisedome assistant architecture.

Different agent frameworks may require different action definitions or route structures.

Performance depends on the similarity between deployment workflows and those represented in the dataset.

Downloads last month
21

Models trained or fine-tuned on bbidpa/wisedome-router-v1-dataset