Etchblok
Etchblok provides self-updating visual maps and AI-generated technical documentation for Python codebases, grounded in a deterministic call-graph analysis rather than pattern matching. The platform automatically opens pull requests with doc updates when code changes merge, integrating into existing GitHub workflows. It targets fast-moving engineering teams that need documentation accuracy to keep pace with AI-assisted development.
- Developer Tools
- Artificial Intelligence
- AI Agents
- Software Only
Funding
Founders
Founder details are not available yet.
Product
Problem
Documentation produced by hand cannot keep pace with code that moves at machine speed, especially for teams shipping with AI assistance. This gap shows up in lost prospects at evaluation, support tickets routed to engineering, and silent churn, while AI agents increasingly consume tech docs as ground truth infrastructure where accuracy is a prerequisite.
Solution
Etchblok builds a deterministic map of your codebase first, then uses AI agents to generate and update technical documentation grounded in that actual system structure. The platform supports Python codebases, resolves call graphs to determine what is truly public API surface, and traces cross-file inheritance for accurate references. When code changes are merged, Etchblok's agents detect the update and open a PR with doc changes, with an option for auto-merge, using existing GitHub workflows. The system also produces living Mermaid diagrams and narrative architecture guides that track actual system behavior as it evolves, eliminating manual diagram maintenance.
Target Audience
Primary customers are fast-moving Series A–C companies with complex Python codebases and high-stakes architectures where documentation accuracy is non-negotiable, such as open source SDK maintainers and platform engineering teams.
Features
- Deterministic codebase mapping that grounds all generated documentation in actual code structure rather than training-data pattern matching
- Automatic PR-based doc updates when code changes are merged, integrating into existing GitHub workflows with optional auto-merge
- Zero-drift re-tracing of the call graph from every changed node outward, catching downstream docs affected by changes across multiple files
- API references that resolve the call graph to identify truly public surfaces, tracing cross-file inheritance and excluding internal methods and deprecated endpoints
- Architecture documentation with living Mermaid diagrams and narrative guides generated from the actual dependency graph
- Full editorial control via GitHub or built-in editor, with review, refine, and publish capabilities