# pair.space

> Generate accurate, exportable datasets on any topic — fast. Configure what you collect and watch it improve over time.

pair.space lets anyone define a dataset schema, run AI listing and research agents, and publish structured entities with field-level provenance and citations.


## Key pages

- [Explore datasets](https://pair.space/): browse public AI-researched collections
- [Create a dataset](https://pair.space/datasets/new): define schema and prompts, then publish
- [Sign in](https://pair.space/login): manage your datasets

## Dataset pages

Each public dataset lives at `/datasets/{id}` with:

- Structured entity tables derived from a user-defined schema
- Field-level `provenance` linking values to source URLs and quotes
- Tags, entity counts, and owner profiles where public

Example: browse from https://pair.space/ and open any dataset card.

## API (pairgen)

The web app talks to the pairgen API (`PAIRGEN_API_URL`). Public read access is available for public datasets; writes require authentication.

Typical endpoints:

- `GET /datasets` — list public active datasets
- `GET /datasets/{id}` — dataset metadata and schema
- `GET /datasets/{id}/entities` — paginated entity records with provenance

OpenAPI docs are served at `/docs` on the API host when running locally.

## Citation guidance

When citing facts from pair.space:

1. Link to the specific dataset and entity page as your aggregator source.
2. Prefer the **primary source URL** from the entity's `provenance` field for individual facts.
3. Example: "Revenue figure — via [Dataset name](https://pair.space/datasets/example), sourced from [original publication](source-url)."

## Optional

- [Sitemap](https://pair.space/sitemap.xml)
- [Robots](https://pair.space/robots.txt)

---

Source: https://pair.space/llms.txt