ANHAD
CHAWLA
CS @ McMaster. I build AI agents that ship real outcomes — an app on the App Store, systems applying to jobs and taking restaurant orders while I sleep.
CS @ McMaster. I build AI agents that ship real outcomes — an app on the App Store, systems applying to jobs and taking restaurant orders while I sleep.
PROFILE.PNG
I design systems that take messy human intent, map it into executable state machines, and ship outcomes without babysitting. The game is not UI clicks, the game is context, memory, and reliable decision logic.
Hover or focus any technology in the grid to inspect capability context in real time.
ONE SECTION. EVERYTHING.
Shipped > talked about. The highlights — each one is real, deployed, and used.
APP STORE
Built and shipped solo: AI face scans that turn your camera into a skin analyst — personalized routines, progress tracking, and a full automated social-marketing pipeline behind it.
AUTONOMOUS
An AI browser agent that applies to jobs like a human: scrapes a 36k-posting corpus, matches to your profile, fills real ATS forms (Workday, Greenhouse…), submits, and verifies via Gmail receipts — tracked on a live dashboard.
A VS Code extension for multi-agent coding: one Leader plans the work and routes sub-tasks to Codex, Cursor, Claude, and Gemini workers — with file-diff tracking and resumable threads built in.
Mobile × web3 campus platform: QR-triggered quests, avatar progression, and on-chain Solana rewards that turn student engagement into a game loop.
Dating without swiping: voice-first onboarding builds an AI digital twin of each user, and twins simulate compatibility conversations in an "Agent Lounge" before a human match is ever surfaced.
Answers real phone orders end-to-end: multi-state Retell agent on Twilio SIP takes the call, an n8n pipeline validates the order and fires SMS/WhatsApp confirmations — running live in a real restaurant.
Real-time satellite positioning + environmental data: custom map interface with AOI selection, synchronized weather retrieval, and orbital path tracking on a Node.js/Express telemetry backend.
NLP chatbot that handles dietary restrictions and food preferences with TensorFlow + spaCy intent models — cut a real restaurant's manual admin work by ~25%.
VISITORS CAN TOUCH THIS
Every system I build has a feedback loop. Here's yours — play the pads (or your keyboard, A S D F G H J K), stack a loop, make something.