Upstash Vector has several features and enhancements planned for its roadmap. These include: 1. Enhanced Search with Metadata Filtering Upstash Vector currently supports storing and retrieving metadata alongside vectors. However, it doesn’t yet allow filtering based on that metadata. This feature will enable you to refine vector search results even further, unlocking more precise and relevant information retrieval. 2. Improved Availability and Performance with Replication Currently, each index resides in a single replica. This means that if the replica becomes unavailable, so does the index. To address this, Upstash Vector is planning to introduce replication. This will involve creating multiple read replicas of each index, ensuring:
  • High Availability: Even if one replica fails, others will remain accessible, maintaining index availability.
  • Reduced Latency: Replicas can be placed in different regions, bringing the index closer to clients and reducing latency for geographically distributed users.
3. Granular Data Management with Namespaces Namespaces will allow you to divide an index into multiple, isolated partitions. Each namespace will function as a self-contained subset of the index. This feature offers significant benefits:
  • Performance Optimization: You can create more performant indexing by dividing large indexes into smaller, more manageable namespaces.
  • Multi-Tenant Data Isolation: Namespaces provide a mechanism to isolate data belonging to different tenants or applications within a single index.