Notebooks
Runnable Colab notebooks covering semantic search, lexical search, hybrid search, RAG, embeddings, reranking, and data ingestion with Pinecone.
Search
Section titled “Search”Full-text search
Search typed JSON documents using BM25 scoring, Lucene query syntax, dense vector ranking of images.
Semantic search
Implement semantic search over an index with dense vectors to find records that are similar in meaning to a given query.
Lexical search
Implement lexical search over an index of sparse vectors to find records that most exactly match the words or phrases in a query.
Cascading retrieval
Implement cascading retrieval (hybrid search with two indexes) to combine the benefits of semantic and lexical search.
Reranking search results
Use Pinecone's reranking feature to enhance the accuracy of search results.
Retrieval-augmented generation (RAG)
Section titled “Retrieval-augmented generation (RAG)”RAG with hybrid search and Claude
Implement simple retrieval-augmented generation with hybrid search and Anthropic's Claude models
Agentic RAG with Claude
Build an agentic RAG pipeline that uses tools to retrieve data from web search and Pinecone semantic search, then generates responses using Anthropic's Claude models
RAG with LangChain and OpenAI
Learn how RAG can be used with Pinecone to reduce hallucinations, by grounding responses using our own release notes
RAG with cascading retrieval and OpenAI
Investigate research papers by implementing cascading retrieval, then pass results to OpenAI to generate answers
Retrieval Agents with Pinecone Assistant, LangChain and LangGraph
Create a study guide generator using agentic retrieval and the Pinecone Assistant Context API
Miscellaneous
Section titled “Miscellaneous”Import documents from JSONL files in an Amazon S3 bucket into a document index