README.md

---
task_categories:
- question-answering
language:
- en
tags:
- web-search
- company-search
- agentic-search
- information-retrieval
- entity-retrieval
- multi-constraint
size_categories:
- n<1K
configs:
- config_name: search_only
  data_files:
  - split: test
    path: data/test.jsonl
---

# OB Company Websearch

10 multi-constraint company-discovery questions with a frozen,
hand-labelled reference set. The public set accompanies the
[OpenBenchmarks Multi Turn Company Search Benchmark](https://openbenchmarks.com/multi-turn-company-search).

The benchmark holds the research agent and budgets fixed while varying the web
search provider. This release is for the **search-only** condition: the agent can
search and use result snippets but is not given a page-fetch tool.

## Dataset contents

Each question asks for the complete set of companies satisfying three or four
simultaneous constraints. This release contains:

- **10 questions**
- **3 three-constraint questions** and **7 four-constraint questions**
- **103 reference-set memberships** across the questions
- **100 unique canonical companies**

Each row contains:

- `case_key` — stable question identifier.
- `question` — natural-language company-discovery request shown to the agent.
- `family` — broad construction family such as investor, accelerator, or funding.
- `constraint_count` and `constraints` — number and text of the simultaneous conditions.
- `predicates` — normalized axes and labels used to construct the question.
- `census_metadata` — dataset-construction checks and reference-set size.
- `source_metadata` — construction coverage metadata.
- `sort_order` — stable display order.
- `gold` — reference companies, including canonical name, domain, aliases, and notes.

Dataset-level release metadata and the source content hash are retained in
`metadata.json`.

## Evaluation

A fixed agent is asked each question and must return the complete matching
company set. Returned companies are resolved to canonical identities and
compared deterministically with `gold`:

- a reference company returned by the agent is a true positive;
- an extra company is a false positive;
- a missed reference company is a false negative.

Precision, recall, F1, and exact-set accuracy are computed per agent run. Each
provider configuration is run three times on the same locked questions; reported
quality metrics are the mean and sample standard deviation across those three
runs.

This repository contains questions and labels, not vendor responses or benchmark
scores.

## Gold construction

Questions were selected from broad intersections in the source company snapshot.
Candidate labels were reviewed in a DB-free local labeller and finalized only
after every case was marked reviewed and the local gold release was published.

## Reproducibility

The open runner and offline judge are available at
[openbenchmarks-labs/multi-turn-company-search](https://github.com/openbenchmarks-labs/multi-turn-company-search).
The benchmark methodology and current results are published at
[openbenchmarks.com/multi-turn-company-search](https://openbenchmarks.com/multi-turn-company-search).

Release: `ob-company-websearch-v1`  
Schema: `company-research-dataset-v1`  
Gold version: `1`  
Content SHA-256: `sha256:b987888daa5e0516ac7cf18deaba3830496929fa653d2cb68b494bd2945664ab`