# Yard — Raw data warehouse

> The warehouse under the curation.

Yard is the raw warehouse of ingested databases that Kiln curates on top of, 2,000+ integrated databases assembled into a knowledge graph of 60B+ RDF triples. It is the substrate our models train on and our data-gap analysis reads from.

## Key facts
- **2,000+** — integrated databases
- **60B+** — RDF triples in the knowledge graph
- **1** — graph the whole platform reads from

## What Yard does

### One knowledge graph
Thousands of sources land in a single graph structure instead of scattered, incompatible files.

### The training substrate
Our in-house models train on the entire Yard graph, giving them far more signal than public data alone.

### Feeds gap analysis
Gaps reads Yard to decide which endpoints have evidence and which are genuinely empty.

## Works closely with
- [Kiln](https://toxindex.com/platform/kiln.md): Curated database layer
- [Pleiome](https://toxindex.com/platform/pleiome.md): Generative multi-property transformer
- [Gaps](https://toxindex.com/platform/gaps.md): Automated data-gap analysis

## Frequently asked questions

**What is Yard?**

Yard is the raw warehouse of ingested databases Kiln curates on top of, 2,000+ integrated databases assembled into a knowledge graph of 60B+ RDF triples.

**Who uses Yard directly?**

Yard is the substrate the in-house models train on and the graph Gaps reads to tell a real data gap from one already covered elsewhere.

**Is Yard the same as Kiln?**

No. Yard is the raw, comprehensive warehouse; Kiln is the curated, normalized layer built on top of it for querying.

---
Canonical page: https://toxindex.com/platform/yard