Arpit Khatri

I’m an AI- and backend-oriented software engineer currently studying Electronics and Instrumentation at Birla Institute of Technology and Science, Pilani. I enjoy turning complex ideas into reliable systems—whether that’s LLM-powered experiences or high-throughput infrastructure.

In my Summer Internship at Caarya, I developed a personalized AI-powered news application using the Gemini API and LangChain. I also built a web scraping pipeline to extract 100+ daily articles from The Times of India, converting them into vector embeddings for semantic similarity search.

I focused heavily on production-ready backend work, including an OAuth 2.0 authentication system with secure handling of access and refresh tokens for protected routes and session persistence. I also engineered distributed and scalable projects, like a concurrent Redis-like in-memory key-value store in Go with leader–follower replication and sub-millisecond latency.

Across projects like Doclytics and real-time chat, I’ve worked with FastAPI, GCP services, and Next.js to deliver analytics and low-latency communication through RESTful endpoints and WebSockets. I’m especially proud of placing 2nd at the Google Cloud AI Accelerate Hackathon among 2700+ participants.