SwellDB is a data system that enables SQL-based analytical querying over dynamically generated tables. Tables are synthesized in real-time from sources including large language models, existing databases, file formats such as CSV and Parquet, and web search results. It creates an open-world query execution model that generates tables on-demand using prompts and schemas, bridging structured SQL querying with unstructured data retrieval.
- Source Code
- https://github.com/SwellDB/SwellDB[01]
- Developer
- Country of Origin
- US
- Start Year
- 2025
- Project Types
- Academic, Open Source
- Written in
- Python
- License
- MIT License
SwellDB is a data system that enables SQL-based analytical querying over dynamically generated tables. Tables are synthesized in real-time from sources including large language models, existing databases, file formats such as CSV and Parquet, and web search results. It creates an open-world query execution model that generates tables on-demand using prompts and schemas, bridging structured SQL querying with unstructured data retrieval.
History
The project started in 2024 at Cornell University.