Open Source
We believe the future of production AI should be built in the open — transparent, collaborative, and accessible to practitioners everywhere.
Projects We Maintain
Cognitive Agent Architecture (CAA)
A reference architecture for building production-grade AI systems.
CAA defines five core layers — Context, Behavior, Execution, State, Collaboration — plus cross-cutting concerns like Observability and Security. Together they turn LLM reasoning into reliable, auditable execution.
[View on GitHub →]
Production Readiness Blueprint
A community-driven playbook for solving real production AI challenges.
Developed through live problem-solving sessions at the appliedAI Developers Meetup – Munich AI Nexus, where practitioners tackle architectural problems and document working solutions for everyone.
[View on GitHub →]
Philosophy
Most agent frameworks stay in prototype mode.
We’re focused on what it takes to ship AI systems that survive contact with reality — clear separation of concerns, explicit control flow, and observability by design.
The architecture informs the community curriculum.
The community challenges refine the architecture.
Both evolve together.
Get Involved
Using CAA in production? → Open an issue or reach out.
Facing a hard AI challenge? → Submit it to the Production Readiness Blueprint — the best get tackled live.
Want to contribute? → Both repos welcome issues, PRs, and RFCs.
Join the Community
We curate and host the Applied AI Developers Meetup – Munich AI Nexus, a bi-monthly gathering of engineers, researchers, and practitioners working on production-grade AI.
Each session focuses on a concrete architectural challenge drawn from the Production Readiness Blueprint and ends with documented takeaways for the community.
🗓 Next Meetup: Luma
📍 Location: Munich
💬 Interested in presenting? Open an issue or reach out.
