Kunal Mudaliar

I'm building production-ready AI applications at MindsyncX Labs, where I help develop scalable Python REST APIs that integrate machine learning models into enterprise environments.

For a vision-based automated attendance system, I architected a React and FastAPI application that detects and identifies people in real-time video streams. I built its facial-recognition pipeline with SCRFD, ArcFace, Supabase, and decoupled Python microservices for high-availability ONNX inference.

I also engineered MindSyncX, a multi-tenant RAG chatbot SaaS platform using FastAPI, Pinecone Vector DB, MongoDB, and React. I addressed LLM hallucination and vector staleness through an automated synchronization pipeline, and built TOTP-based MFA, RBAC, live session tracking, and token-usage analytics.

I enjoy writing clean, efficient code and bringing AI/ML models into fast, reliable full-stack products.