AI Debugging Agent (Ollama & Pinecone)
Mar 2026- End-to-end agent with FastAPI and Next.js (TypeScript): ingests GitHub/ZIP repos, chunks and embeds source, answers natural-language debugging queries with file/line references.
- Modular RAG pipeline with SentenceTransformers, CodeLlama via Ollama—retrieval orchestration, structured JSON outputs, SQLite query history, optional pylint/eslint for context.
- Production practices: pipeline logging, robust error handling with LLM/Pinecone fallbacks, namespace multi-repo queries, saved-repo dropdown, privacy-first local vector mode.
- FastAPI
- Next.js
- RAG
- Ollama
- Pinecone

