Redis

Redis builds the eponymous in-memory database, originally known as a key-value store and now increasingly positioned as a real-time context engine for AI applications. Redis-powered vector search retrieves structured and unstructured data in real time, so language models always work with up-to-date context.

For AI agents, Redis Agent Memory manages short- and long-term memory so agents stay coherent across multiple interactions, while Redis Data Integration syncs data from existing databases in real time. LangCache adds semantic caching for LLM requests, aimed at noticeably lowering both response times and API costs.

Through the official Redis MCP Server, AI agents can access these capabilities directly without implementing their own Redis clients. The business model combines an open-source core with the commercial Redis Enterprise offering for production-scale deployments; according to the company, 43% of enterprise agent stacks already use Redis to store agent data.