Give your AI agents long-term memory without building the whole memory system yourself.
MouseBase provides a simple API to store, search, inspect, and delete memories for AI agents. It works with Python, JavaScript, or REST and is backed by PostgreSQL and pgvector.
Built for developers who want their agents to remember users, conversations, preferences, and context across sessions.
I’ve been building MouseBase because I got tired of AI agents forgetting everything between conversations.
The idea is pretty simple. Give AI agents a proper long-term memory that developers can actually plug into without having to build the whole thing from scratch.
MouseBase lets you store, search, inspect, and delete memories through a simple API, with PostgreSQL and pgvector underneath.
It’s still early, and I’m building this as a solo developer, so I’d really love to hear what you think.
If you’re building AI agents, how are you handling memory right now? What’s frustrating about your current setup?
Hey everyone!
I’ve been building MouseBase because I got tired of AI agents forgetting everything between conversations.
The idea is pretty simple. Give AI agents a proper long-term memory that developers can actually plug into without having to build the whole thing from scratch.
MouseBase lets you store, search, inspect, and delete memories through a simple API, with PostgreSQL and pgvector underneath.
It’s still early, and I’m building this as a solo developer, so I’d really love to hear what you think.
If you’re building AI agents, how are you handling memory right now? What’s frustrating about your current setup?
Would love to hear your thoughts and feedback.