Welcome to Weaviate Recipes 💚
This repository covers end-to-end examples of the various features and integrations with Weaviate.
| Category | Description |
|---|---|
| Integrations | Notebooks showing you how to use Weaviate plus another technology |
| Weaviate Features | Notebooks covering vector, hybrid and generative search, reranking, multi-tenancy and more |
Integrations 🌐
Check out Weaviate's Integrations Documentation!
| Company Category | Companies |
|---|---|
| Cloud Hyperscalers | Google, AWS, NVIDIA |
| Compute Infrastructure | Modal, Replicate |
| Data Platforms | Databricks, Confluent Cloud, Spark, Unstructured, Firecrawl, Context Data, Aryn, Astronomer, Airbyte |
| LLM Frameworks | DSPy, LangChain, LlamaIndex, Semantic Kernel, Ollama, Composio, Haystack |
| Operations | Arize, Langtrace, LangWatch, Nomic, Ragas, Weights & Biases |
Weaviate Features 🔧
| Feature | Description |
|---|---|
| Similarity Search | Use Weaviate's nearText operator to run semantic search queries (broken out by model provider) |
| Hybrid Search | Use Weaviate's hybrid operator to run hybrid search queries (broken out by model provider) |
| Generative Search | Build a simple RAG workflow using Weaviate's .generate (broken out by model provider) |
| Filters | Narrow down your search results by adding filters to your queries |
| Reranking | Add reranking to your pipeline to improve search results (broken out by model provider) |
| Media Search | Use Weaviate's nearImage and nearVideo operator to search using images and videos |
| Classification | Learn how to use KNN and zero-shot classification |
| Multi-Tenancy | Store tenants on separate shards for complete data isolation |
| Product Quantization | Compress vector embeddings and reduce the memory footprint using Weaviate's PQ feature |
| Evaluation | Evaluate your search system |
| CRUD APIs | Learn how to use Weaviate's Create, Read, Update, and Delete APIs |
| Generative Feedback Loops | Write back to your database by storing the language model outputs |
Feedback ❓
Please note this is an ongoing project, and updates will be made frequently. If you have a feature you would like to see, please create a GitHub issue or feel free to contribute one yourself!