I'm a Computer Science and Artificial Intelligence undergraduate at IIIT Delhi, interested in the space between AI, solid software engineering, and products people can actually use.

I enjoy exploring an idea from multiple angles: the user problem, the algorithms underneath, the architecture, and the failure cases. Lately, a lot of my curiosity has been around AI-powered learning tools, developer workflows, and making AI systems more transparent and reliable.

• 🧠 Applied AI: machine learning, RAG, learning tools, and transformer tooling
• 🛠️ Developer tools: useful CLIs, automation, and workflows that make development easier to reason about
• 🌐 Product engineering: taking ideas from interface to backend
• 📚 CS foundations: DSA, graph algorithms, databases, and systems concepts

A CLI for analyzing repositories and managing transformer-optimization patches. It supports dry runs, reversible changes, and validation that distinguishes pass, fail, and skipped instead of claiming work ran when it didn't.

An AI-powered study companion that turns documents into interactive learning: document-grounded chat, flashcards, quizzes, and summaries.

A privacy-first browser extension designed to help people filter low-quality feed content and AI-style articles, with a focus on local processing and user control.

An experimental decision-security gateway for inspecting untrusted AI-agent inputs and proposed tool calls, with explicit policy checks and redacted audit records. It's a prototype not a production security boundary.

A Python/Flask route planner exploring graph algorithms through a practical problem: finding routes by distance or line changes, with fare and travel-time estimates.

More projects: SEO Agent · Medical Reception Chatbot · Hate Speech Detection

Languages: Python · Java · JavaScript · TypeScript · SQL
AI & data: NumPy · pandas · scikit-learn · RAG and LLM workflows
Web & backend: Flask · FastAPI · React · Next.js
Workflow: Git · GitHub · Linux · VS Code

I'm continuing to deepen my foundations in DSA, machine learning, Java backend development, and systems internals. The goal isn't to collect tools it's to understand when and why to use them.

• Start with the problem. A project should solve something concrete, not just wear a trendy tech stack.
• Make behavior explainable. Clear outputs and honest validation beat impressive-looking claims.
• Think about failure cases. Privacy, safety, and predictable behavior belong in the design—not as afterthoughts.
• Keep improving. Small iterations, better fundamentals, and a little kaizen every day.

You'll probably find me on a basketball court 🏀, playing football ⚽, tuning a keyboard ⌨️ and other PC peripherals, making coffee ☕, or watching anime and sports.