recipes

Why we included this project

Weaviate teams that are tired of reading docs and guessing at the wiring will find a working starting point in this collection of runnable Jupyter notebooks. Each one walks through a concrete feature end to end, from plain vector and hybrid search to reranking, generative output, media search, and multi-tenant setups, all against a ready dataset. The integrations section pairs Weaviate with the frameworks you likely already have in the stack, such as LangChain, LlamaIndex, DSPy, and Haystack, along with data, evaluation, and observability tools, so you can see how the pieces connect before building your own pipeline. Notebooks are self-contained, which makes the repository work both as a quick reference for a single operation and as a structured learning path for engineers newer to semantic search.