build β ship β repeat β¨
watch the little contribution snake collect your coding days β¨
|
|
An AI-powered codebase explorer that connects GitHub repositories, indexes code with embeddings, and lets developers ask natural-language questions with grounded answers and real file/line citations. What makes it fun: structure-aware chunking, Gemini embeddings, MongoDB Atlas Vector Search, GitHub OAuth, a BullMQ/Redis re-indexing queue, and a 3D dependency graph β all built around one rule: answers should trace back to retrievable code.
|
π DevTinderThe frontend for a full-stack developer networking platform: discover people by skills, connect, chat in real time, and explore GitHub-verified profiles.
|
πΈ SnapSpendAn AI expense tracker that turns receipt photos into structured expenses, answers questions about real Firestore data, and keeps budgeting human-friendly.
|
π¬ NetflixGPTA Netflix-inspired movie discovery app with streamed Gemini recommendations, function calling, multimodal image search, TMDB data, and real-time Firestore-backed lists.
|
π οΈ CodeAtlasThe project Iβm pushing deepest right now: a RAG assistant for real repositories, with evaluation harnesses, citations, semantic search, and background re-indexing.
|
| βοΈ TypeScript depth | βοΈ Cloud & DevOps | βοΈ Backend architecture | βοΈ RAG evaluation | βοΈ AI application engineering |
thanks for visiting Β· making my dreams a commit at a time β¨



