How GeraMind works
Memory and context layer for AI agents — yours, portable, private.
Quick answers
- What is GeraMind?
- A persistent memory layer for AI agents. It stores your facts, preferences, summaries, and embeddings under your control, and exposes them to compatible agents via a typed schema.
- Why do I need a separate memory layer?
- Without one, your context is locked inside whichever assistant you used last. With GeraMind, switching to a new model or agent doesn't mean starting over — your memory follows you.
- Which agents can read my memory?
- Only the ones you authorise, with the scopes you grant. Most major assistants and any Nexus-compatible agent are supported. You see and revoke access at any time.
- How is my memory stored?
- Encrypted at rest in your private workspace, EU region by default. Embeddings are computed on a model you pick. We do not use your memory to train any model.
The journey, step by step
- 1
Set up your memory
Create a private GeraMind workspace. Add facts (your preferences, projects, contacts) manually or import from existing notes.
- 2
Connect your agents
Authorise the agents you trust — ChatGPT, Claude, GeraOS, your own. They read and write under the scopes you grant.
- 3
Use anywhere
Switch assistants without losing context. Audit every read and write. Revoke or export at any time.
Ready to start?
GeraMind is the persistent memory layer for AI agents. It stores facts, preferences, and history under user control, exposes them to any compatible agent via a typed schema, and never locks your conte