Varun chandra Kola

I’m an M.Tech Computer Science student with hands-on experience building and deploying AI systems in production. I focus on practical ML—turning models into reliable services with clean APIs, scalable deployment, and measurable performance improvements.

I delivered a multilingual RAG-based medical chatbot (Medi-Assist AI), integrating Pinecone vector embeddings and HuggingFace/Google Gemini for context-aware responses. I engineered a secure Flask backend with JWT authentication, MongoDB session management, and Docker, then deployed it on Google Cloud Run—reducing average query latency to sub-2 seconds across multilingual inputs.

I also built an ensemble deep learning classifier for preliminary dermatological diagnosis, combining MobileNetV2, Capsule Networks, and a non-linear SVM to classify 12 skin disease categories. In a national hackathon (DAET), I served as the sole technical contributor, developing a FastAPI + Uvicorn backend integrated with Supabase (PostgreSQL + object storage) featuring SHA-256 duplicate detection, automated schema inference, and large-scale dataset profiling.

Beyond core AI work, I’ve shipped full-stack features via a MERN internship—owning a community module end-to-end and improving engagement through threaded replies and notification hooks. I’m expanding into MLOps engineering with MLflow, GitHub Actions CI/CD, and AWS, and I’m seeking entry-level MLOps / ML Engineer roles to build scalable, automated ML pipelines.