Feast v0.65 ships a native ScyllaDB online store with vector search support. Which means you can store your ML features alongside vector embeddings in the same database and query both using Feast.
What’s new:
- Shard-aware Python driver
- Self-hosted or ScyllaDB Cloud
- Implements Feast vector database API
- Query features and embeddings using the same interface
- Performance-optimized schema
Read more: Feast Feature and Vector Serving at ScyllaDB Speed
Docs to get started: Feast | ScyllaDB Docs