
Observable memory and caching for your AI agents
Open-core memory and caching for AI agents, on your own Valkey or Redis. Unlike a black-box hosted cache, you see every decision and can change it while the app runs. Semantic, multi-tier caching answers repeat LLM calls in under 1ms; memory gives scoped, ranked recall. The cache tunes itself: an MCP agent proposes a threshold or TTL change, a human approves, and the app picks it up live, no redeploy. No other Valkey or Redis cache does this. MIT, self-hostable. Demo: chat.betterdb.com

Open-core memory and caching for AI agents, on your own Valkey or Redis. Unlike a black-box hosted cache, you see every decision and can change it while the app runs. Semantic, multi-tier caching answers repeat LLM calls in under 1ms; memory gives scoped, ranked recall. The cache tunes itself: an MCP agent proposes a threshold or TTL change, a human approves, and the app picks it up live, no redeploy. No other Valkey or Redis cache does this. MIT, self-hostable. Demo: chat.betterdb.com
1 comment
Memory and caching for your agents. Self-host on your own Valkey or run it managed, either way you see every decision and can change it live. No black box. Live at chat.betterdb.com, feedback welcome.