Deep Sharma

I’m a full-stack developer with internship experience building MERN applications, and hands-on work across TypeScript, PostgreSQL, Docker, and AI/RAG systems. I focus on shipping features end-to-end, from API design to user-facing interfaces.

In my Software Developer Intern role, I built REST APIs and React interfaces for a doctor appointment platform, covering appointment booking, cancellation, and status management. I structured the backend using MVC architecture with routes, controllers, models, and middleware across React.js, Node.js, Express.js, and MongoDB, and implemented JWT-based authentication with role-based access control.

My projects reflect my passion for applied AI: SupportFlow AI delivers token-by-token streaming LLM responses using Server-Sent Events, plus persistent conversation and message history stored via PostgreSQL and Prisma. I also added request validation, centralized error handling, and structured logging, and containerized the workflow with Docker for consistent setup.

With DocuQuery, I built a retrieval-augmented generation pipeline for multi-PDF semantic Q&A using LangChain, FAISS, Groq, and Hugging Face embeddings, returning grounded answers with source filenames and page numbers. I enjoy turning real constraints (like vector-store issues) into better architecture, and building interactive experiences with Streamlit.