datachain-ai/datachain: The Context Layer for unstructured data: typed, versioned datasets over S3, GCS, Azure
The Context Layer for unstructured data: typed, versioned datasets over S3, GCS, Azure - datachain-ai/datachain
Open Source AI Tool
Summary
Dash is an open‑source self‑learning data agent that generates SQL queries using six layers of contextual information, from schema metadata to runtime error learning. It continuously improves by storing successful query patterns and error corrections without requiring model retraining.
Highlights
- Six‑layer context hierarchy (schema, annotations, patterns, institutional knowledge, error learnings, runtime schema)
- Self‑learning loop that records successful queries and fixes errors automatically
- Zero‑retraining approach; improvements are stored as JSON/SQL knowledge files
- Easy deployment via Docker with sample datasets such as the F1 dataset
- Extensible: users can add custom knowledge bases in JSON or SQL
auto-generated
Context
Audience
Developers building AI‑augmented data pipelines and SQL assistants, data engineers interested in self‑learning agents, and teams looking for an extensible open‑source alternative to proprietary solutions
DomainData Engineering and AI
Formatopen‑source software repository (GitHub)
Accessfree open source on GitHub, Docker deployment
Topics
Related
LangChainOpenAI Function CallingSQLChatPostgreSQLData Agent patterns
