LOTUS is an open-source Python query engine using LLMs and a Pandas-like API for fast document processing. It optimizes queries for up to 400x speedups and supports declarative AI pipelines for tasks like fact-checking and search ranking.
Highlights
Pandas-like API with LLM-powered semantic operators
Up to 400x performance optimization
Declarative multi-step AI pipeline support
State-of-the-art accuracy on classification tasks
Seamless integration with vector databases
auto-generated
LOTUS Team at Stanford and Berkeley University · via LOTUS
PandasLarge Language ModelsSemantic SearchDocument ProcessingVector DatabasesFEVER Dataset
Discover Similar Content
parallel.ai
Parallel Web Systems | Enterprise Deep Research API
Enterprise deep research API with structured data extraction for AI agents. Up to 58% accuracy outperforming GPT-5 (41%), Exa (14%), Anthropic (7%), P...
github.com
GitHub - joelhooks/pdf-library: Local PDF knowledge base with vector search using PGlite + pgvector
Local PDF knowledge base with vector search using PGlite + pgvector - joelhooks/pdf-library
softwaredoug.com
Turning my laptop into a Search Relevance Judge with local LLMs
Local LLMs can evaluate 100s of result pairs a minute in a Macbook, enabling a new age of rapid search relevance improvements