Naman Gupta

About Naman Gupta

I’m a final-year B.Tech student in AI & ML at Oriental Institute of Science and Technology, focused on becoming a top-tier AI engineer who ships real systems not just prototypes. I build production-oriented AI solutions across document intelligence, retrieval pipelines, agentic workflows, analytics, and developer tooling.Currently, I’m working as an AI Engineer Intern at Infradock AI, researching and prototyping intelligent systems that make production environments faster, more reliable, and more autonomous.Previously, I’ve delivered end-to-end AI and backend systems across multiple teams:• Built Retrieval-Augmented Generation (RAG) pipelines using FastAPI, LangChain, and ChromaDB for contextual AI features in full-stack apps• Integrated modern Next.js frontends and deployed via Docker + AWS CI/CD, reducing rollback time by 60%• Shipped 10+ secure REST APIs with Node.js/Express and developed a TypeScript React component library that improved development speed by 35%• Designed scalable microservices with MongoDB, lowering failure rates by 30% through strict validation and reliability practicesI also build independent products and tools:• FinTrack — MERN personal finance platform with real-time ML insights• CampusMitra — RAG-based AI assistant fine-tuned for domain accuracy using Hugging Face + ChromaDB• Ongoing work on AI agents, automation tools, and production-ready FastAPI/Next.js stacksMy interests lie in applied AI systems that combine LLMs, data pipelines, backend engineering, and product thinking to solve real problems at scale.Currently open to:• Global AI / Software Engineering internships• Startup collaborations and founding-engineer roles• Freelance AI and backend projects• Open-source and research collaborationsIf you’re building something ambitious with AI, distributed systems, or developer platforms let’s connect.

Experience

1. Ai Engineer Intern

   Infradock Ai

   Jan 2026 — Present · Delhi, IN

   Engineered multi-agent workflows using LangGraph and LangChain, automating resolution for 500+ dailyqueries and reducing manual review time by ∼60%.– Optimized RAG pipeline accuracy by ∼40%(achieving 95%+ precision) through advanced semantic chunking,clause-level boundary detection, and cross-encoder reranking via Qdrant.– Architected scalable document processing pipelines using Cloudflare R2, processing documents, accel-erating data ingestion by 4x, and reducing retrieval latency by 30%

Education

• Oriental Institute of Science & Technology

  Bachelor of Technology - BTech, Artificial Intelligence and Machine learning

  2026

• St. Joseph's Co-Ed School

  High School Diploma, Mathematics and Computer Science