Finding Birds You Can't Name with Pinecone Full Text Search

Discover how Pinecone's new full-text search capabilities let you combine text and vector search in a single index. In this demo, we build Bird Search — a clever multimodal app that searches 2,000 Wikipedia bird articles using exact phrase matching, Lucene query syntax, and visual search powered by Gemini Embedding 2 vectors generated from bird images. See how pre-filtering with text queries supercharges vector search to help you identify birds by description, location, and behavior — all from one Pinecone index. Resources: 🔗 Bird Search Demo Repo: https://github.com/pinecone-io/bird-search-example 🔗 Pinecone Full-Text Search Blog: https://www.pinecone.io/blog/full-text-search-architecture/ 🔗 Gemini Embedding 2 Docs: https://ai.google.dev/gemini-api/docs/embeddings Follow Pinecone: https://twitter.com/pinecone https://www.linkedin.com/company/pinecone-io/ https://discord.gg/qwQPPPZ4nz 0:00 Intro to Pinecone full-text search 0:24 Bird Search app overview 0:38 Exact phrase match search 1:05 Lucene query syntax with boosting 1:25 Visual search with Gemini Embedding 2 1:55 Combined text and vector search 2:28 Pre-filtering with full-text search 3:10 Results and how pre-filtering works 3:25 Start building with Pinecone full-text search #Pinecone #FullTextSearch #VectorSearch #MultimodalSearch #GeminiEmbedding