This browser cannot draw the portrait (WebGL is off).
The biggest bottleneck in AI
Structuring and maintaining information that agents can find and understand.
AI needs context
You should not have to explain the whole story every time you ask an agent to act.
Context needs structure
The information exists. The next problem is making it possible to find.
Prune and maintain
Keep what matters. Update what changes. Remove what no longer helps.
Bernat Sampera
Jev finds prepared answers about me and my work.
How does a company become AI native?
How should AI store and find information?
Recent writing
Ideas and decisions from the work.
How we built the MAAT Admin Agent for gym owners
We built an agent inside the MAAT Admin Dashboard. It reads the backend API through MCP-style actions, uses Jev to plan each turn, shows widgets in the chat, and is tested with a bench and real conversations.
Global context for AI agents: minimal to load, fast to find
An AI model does not know your business. A global context is a map of that information, split into domains. Each session loads only the domain the question needs, and saves what changed.
Building an AI Brain at MAAT
We are building context that people and AI can follow. A clear hierarchy connects company knowledge to reviewed procedures, so more of the team can use it. Adding knowledge as it grows is the next challenge.
