Welcome to the CHEN documentation site. CHEN is a distributed inference architecture that replaces a single monolithic hyper-scale model with a coordinated garage of specialized, low-parameter models.
- Architecture — system diagrams (Mermaid) and component overview
- Math specs — formal definitions of EPU, cost model, latent nuance score
- ADRs — architecture decision records (why each choice was made)
- Operations — deployment, observability, runbooks
- Threat model — security analysis
- Governance — how decisions are made
See the README on GitHub for installation and quick-start instructions.
Auto-generated API docs are available via mkdocstrings once you run
mkdocs serve locally. They cover every public class and function in
src/chen/.