I build and ship AI products that actually solve a problem. Not just the demos that sits in github repo :)
In 2021, I typed something into the OpenAI Playground. Got a response in seconds. Couldn't stop thinking about it.
Spent a year doing client web work — decent money, zero excitement. Kept coming back to AI every evening. Testing, breaking, rebuilding.
At some point the pretending stopped. I stopped treating AI as a side interest and went all in.
I care about the layer where research meets production — real data, real edge cases, real users.
Every project I've built started with a real frustration. Minutz because meeting bots are annoying. LegalMind because contracts shouldn't need a lawyer to read.
An invisible, browser-native meeting recorder that intercepts WebRTC audio streams directly inside Chrome with a full FastAPI & Next.js pipeline.
Training a 24-class SwinV2 classifier on 28K+ dermatological images achieving 78% accuracy, combined with dermatologist discovery.
A shipped Chrome extension for 1-click summarisation, entity extraction, and contextual Q&A with multi-provider LLM abstraction.
Building a production RAG pipeline for legal documents with 12 parallel async LLM risk-extraction tasks targeting under 30s latency on 50 pages.
Infosys Springboard
Sep 2025 – Nov 2025
Contra Platform
Oct 2023 – Feb 2024
Mumbai, India
7.00/10 CGPA
Graduated: 2026
Mumbai, India
3rd Rank in Department
Graduated: 2023
“When you're winning, you're not as good as you think you're.
When you're losing, you're not as bad as you think you're.”
