Description
Weaviate Recipes (Python) is a GitHub repository of end-to-end Jupyter notebook examples demonstrating how to use Weaviate as a vector database from Python. The notebooks cover data ingestion, vector and hybrid search, generative AI workflows, RAG (Retrieval-Augmented Generation), and integrations with a wide range of tools and platforms.
Features
Repository Structure
-
Datasets
- Ready-to-use datasets for ingestion into a Weaviate cluster.
- Supports examples of loading and indexing data for search and RAG scenarios.
-
Integrations
- Notebooks showing how to use Weaviate alongside other technologies and platforms.
- Focus on combining Weaviate’s vector search with external compute, LLMs, data pipelines, and observability tools.
-
Weaviate Features
- Notebooks demonstrating Weaviate core capabilities, including:
- Vector search
- Hybrid search
- Generative search
- Reranking
- Multi-tenancy
- Use of
nearText, hybrid, and .generate operators with multiple model providers
-
Weaviate Services
- Examples of how to build applications using Weaviate Services.
- End-to-end workflows showing service-oriented usage patterns.
Integrations Coverage
The repository aligns with Weaviate’s broader integrations ecosystem; example notebooks focus on connecting to tools from categories such as:
These integrations are reflected in notebooks that show how to combine Weaviate with LLM frameworks, data ingestion pipelines, evaluation tools, and MLOps platforms.
Weaviate Feature Examples
Some of the Weaviate functionality illustrated in the notebooks includes:
- Using model providers with Weaviate’s
nearText, hybrid, and .generate operators.
- Applying filters to narrow search results (where-clause style queries, structured filtering alongside vector search).
- Implementing vector search for semantic retrieval.
- Implementing hybrid search that mixes keyword and vector-based retrieval.
- Using generative search to produce answers or summaries from retrieved context.
- Demonstrating reranking strategies to improve result ordering.
- Showcasing multi-tenancy patterns where applicable.
Use Cases
- Learning how to connect a Python application to a Weaviate cluster.
- Building RAG pipelines that combine Weaviate with LLM and agent frameworks.
- Prototyping vector and hybrid search over your own or example datasets.
- Exploring integrations with cloud providers, data platforms, and evaluation tools.
- Using complete, runnable Jupyter notebooks as templates for production systems.
Pricing
- The Weaviate Recipes (Python) repository itself is an open GitHub resource; no pricing information is specified in the provided content.
- Any costs would come from underlying platforms or services used in the examples (e.g., cloud providers, external model or data services), which are not detailed here.
