treygrainger/ai-powered-search

Implements semantic search, retrieval-augmented generation, learning-to-rank, and personalized search using dense vector embeddings, LLMs, and user behavioral signals. All examples are Jupyter notebooks in Python with PySpark for data processing, abstracted across multiple search engines (Apache Solr, Elasticsearch, Weaviate, Pinecone, and others) via pluggable engine implementations. Runs entirely in Docker containers for simplified environment setup and reproducibility.

372 stars.