Building the platform for AI systems your whole team can trust.

Our thesis

As AI systems grow, business rules, prompts, and execution details can become harder to trace and maintain. A policy change, revised tone, or new regulatory constraint should have a clear place to live, a clear owner, and a reviewable history.

Orchetree structures AI systems as agent trees, where engineering owns the architecture and business teams can own the domain logic. A tree gives every responsibility a node, every prompt block a role, and every change a clear owner. The result is an AI system the whole organization can inspect, govern, and evolve together.

Open source and self-hosted

The full Orchetree editor, Co-pilot, SDK, backend APIs, MCP endpoint, and portable exports are available under Apache-2.0. You can inspect and adapt the source, deploy it in your own or a client's environment, build commercial solutions, and operate hosted services. The Orchetree name and logos identify the official project and services.

Contact

The best way to reach the project is by email or through the GitHub repository.