LobeEngineAI
- Research
- 600+ automated tests in Go
Research and development: a cognitive operating system that organises AI agents as lobes of a brain, with auditable memory, verification and learning.
- Applied AI
The problem
AI agents answer confidently even without evidence, forget what they learned and never show why they decided.
What the system does
A Go kernel with deterministic engines: memory with validation labels, adaptive reasoning as a step plan, verification against criteria fixed before execution, experience learning compared on a benchmark, and a 3D observatory that shows every lobe at work.
System screens
The screens stay in Portuguese, as in the system.
Modules
- Cortex and specialised lobes
- Neural bus (NATS)
- Cognitive memory
- Adaptive reasoning
- Verification
- Experience learning
- World model and simulation
- Brain Observatory (WebGL)
What the screen is made of
Every screen above depends on decisions like these, made in the code and covered by tests.
| Part | Decision |
|---|---|
| Verification | Acceptance criteria fixed before execution; without evidence, the verifier fails the task. |
| Learning | A candidate policy is only promoted after being compared with the active one on the benchmark. |
| Tools | APIs authorised by name, no redirects, and writes only to a draft area. |
Technology
- Go
- TypeScript
- WebGL
- NATS
- Docker
Stage
Research
Internal research project. The captures come from the development environment with no language model configured, which is why the verifier is shown failing anything without evidence.
Want to see this system working?
We give a guided demo with demonstration data and talk about what it would solve in your operation.
Request a demo