About
A bit about me.
I'm a Computer Science student at Purdue University (Class of 2029) with a minor in Artificial Intelligence. I grew up in Belgium as one of very few Indian students at my school, then moved to Downingtown STEM Academy in Pennsylvania before landing in West Lafayette.
I build full-stack ML systems focused on production-grade engineering — real test coverage, real deployments, real users. From PyTorch classifiers trained on thousands of commits to REST APIs serving 400+ participants, I care about things that actually work in the wild, not just in notebooks.
When I'm not building, I'm at the poker table, the ping pong table, or the pool table. The three P's.
—Google Developer Groups·SWE·Purdue Personal Trainer
—ML@Purdue·ML Engineer·CodeSentry
—Hack the Future·SWE·Pratigya Tournament Platform
DEGREE
B.S. Computer Science · Minor in AI
UNIVERSITY
Purdue University, West Lafayette, IN
GRADUATING
May 2029
BACKGROUND
Belgium → Downingtown, PA → West Lafayette
Experience
Where I've worked.
ML Engineer — CodeSentry
ML@Purdue · West Lafayette, IN
Mar 2026 – Present
- Built an AI code review bot deployed as a GitHub App that auto-reviews PRs with cross-file context awareness, analyzing function signatures, PR metadata, and commit history across all changed files to catch bugs spanning multiple modules
- Trained a PyTorch bug-risk classifier on 1,200+ real commits from 15 repos (Django, React, Flask) with 5 semantic features, achieving 73.4% recall at 2ms per PR
- Engineered a two-stage triage pipeline: classifier scores chunks and enriches with semantic adjustments, only high-risk code goes to Claude with file structure and risk focus areas — cutting API costs 40–60% while adding confidence filtering and duplicate grouping
- Shipped with 62 passing tests, a React dashboard with real-time snippet scanning, and inline PR comments with fix suggestions and severity labels. Deployed on Render with HMAC webhook verification
Oct 2025 – Present
SWE — Pratigya Tournament Platform
Hack the Future · West Lafayette, IN
- Architected full REST API backend serving 400+ participants across 32 teams
- Role-based access control middleware with JWT authentication across 15+ protected endpoints
- Engineered automated eligibility pipeline computing player and team status from document verification, team size constraints, and compliance across 6 Mongoose schemas
- Developed Haversine-formula geolocation API to rank teams by proximity for intelligent match scheduling
Jan 2026 – Present
SWE — Purdue Personal Trainer
Google Developer Groups · West Lafayette, IN
- Built end-to-end async pipeline: Riverpod → Dio HTTP → Firebase Cloud Functions → Gemini API
- Reactive plan generation with Firestore document caching for offline support
- Polished chat UI with themed message bubbles and animated typing indicator using phased AnimationController
Nov 2023 – Aug 2025
Coding Instructor
Code Ninjas · Downingtown, PA
- Designed and delivered project-based curriculum in Java, Python, and JavaScript for 70+ students, providing 1-on-1 debugging support and mentorship from fundamentals through full project completion
- Drove 17+ new enrollments through improved student engagement and structured lesson planning, contributing directly to center revenue growth
Projects
What I've built.
── ◆ ──
01MAR 2026
CodeSentry
AI-powered code review GitHub App. PyTorch classifier + Claude API for intelligent PR review.
- ◆Trained a PyTorch bug-risk classifier on 1,200+ real commits from 15 repos (Django, React, Flask) with 5 semantic features, achieving 73.4% recall at 2ms per PR
- ◆Engineered a two-stage triage pipeline: classifier scores chunks and enriches with semantic adjustments, only high-risk code goes to Claude — cutting API costs 40–60%
- ◆Shipped with 62 passing tests, React dashboard with real-time snippet scanning, and inline PR comments with fix suggestions and severity labels
- ◆Deployed on Render with HMAC webhook verification, CORS lockdown, and rate limiting
02MAR 2026
PokerLens
Full-stack poker equity calculator replacing $250+ commercial tools like PioSolver.
- ◆Trained a PyTorch neural network on 200K Monte Carlo simulations achieving 94.1% correlation, reducing equity inference from 500ms to 0.08ms — a 6,000x speedup
- ◆Built and deployed 6 features across a React frontend (Vercel) and FastAPI backend (Render) with Google/GitHub OAuth via Supabase
- ◆Engineered an action advisor computing pot odds, drawing outs, and fold/call/raise recommendations with step-by-step mathematical reasoning per decision
- ◆CORS lockdown, per-IP rate limiting (30 req/min), HMAC webhook verification, Supabase row-level security
03SEP 2023 – JUN 2024
SymptoScan
AI diagnostic tool analyzing MIMIC-IV medical data. 5th nationally at TSA.
- ◆Analyzed 40,000+ medical cases from MIMIC-IV, achieving 12% reduction in diagnostic error rate through probabilistic symptom-to-diagnosis matching
- ◆Placed 5th nationally at TSA National Conference for Software Development after earning 1st at State
PythonReact
042023
LegalLens
AI-powered mobile app simplifying complex legal documents for lower-literacy individuals.
- ◆Translates dense legal language into accessible text for individuals with lower literacy levels
- ◆Certificate of Special Congressional Recognition from PA-06 representative
AI/MLMobile Development
Education
Purdue University
B.S. Computer Science · Minor in Artificial Intelligence
West Lafayette, IN · Expected May 2029
Relevant Coursework
Data Structures (CS 25100)Computer Architecture (CS 25000)Programming in C (CS 24000)Foundations of CS (CS 18200)OOP (CS 18000)Multivariate Calculus (MA 26100)
Recognition
Honors & Awards.
5th
TSA National Conference — Software Development
June 2024 · National
Top 12
TSA National Conference — Data Science & Analytics
June 2024 · National
1st
PA TSA States — Data Science & Analytics
April 2024 · State
1st
PA TSA States — Software Development
April 2024 · State
2nd
Congressional App Challenge PA-06 — LegalLens
January 2024 · District
Skills
What I work with.
Languages
PythonJavaScriptTypeScriptJavaCDartSQLHTML/CSS
ML & Data
PyTorchNumPypandasspaCyHugging FaceMatplotlib
Frontend
ReactTailwind CSSViteFlutterRiverpod
Backend
FastAPINode.jsExpress.jsMongooseFirebase Functions
Cloud & Databases
SupabaseFirebaseMongoDBGoogle CloudVercelRender
Tools
GitDockerPostmanFigmaVS CodeIntelliJ
Get in touch
Let's talk.
Open to conversations about engineering, AI, or just a good problem to solve.
Open to SWE/ML internships for Summer 2027