
Modern enterprises inherit massive, undocumented codebases that are costly to maintain and hard for new developers to ramp up on. AI Software Archaeologist solves this pain point by combining static analysis, a PostgreSQL‑backed knowledge graph, and large‑language‑model (LLM) reasoning. The backend (Node / Express + Prisma) runs on Render, while a sleek React + Vite frontend is hosted on Vercel. Users authenticate via JWT, then can import any public GitHub repository. The platform instantly produces: • Interactive Architecture Maps (module‑dependency graphs). • Technical‑Debt Scores (complexity, duplication, security, maintainability, etc.). • AI Chat & Documentation – ask natural‑language questions and get auto‑generated markdown docs. • Onboarding Roadmaps – step‑by‑step learning paths for new team members. All data is stored in a managed PostgreSQL instance, making the service fully cloud‑native, scalable, and ready for production use.
17 May 2026