Building intelligent systems at the intersection of theory and practice.
I'm a junior at Tulane University studying Computer Science and Mathematics, with a minor in AI and Business. I'm here on a South Asian Scholar full-tuition merit award, carry a 3.95 GPA, and have made the Dean's List every semester. Graduating May 2028.
I'm a research assistant at the Tulane Data and AI Collective (TUDAI), where I work on multi-agent systems and social choice theory — we co-authored a paper accepted at AAMAS 2026. I also TA two CS courses (CMPS3160 Data Science and CMPS1500 Intro to CS), running labs and office hours for 50+ students.
Outside of code I'm a PADI Master Scuba Diver, Rescue Diver, and underwater videographer. I think a lot about how intelligent systems and ocean conservation can meet in the middle. Also: Formula 1, Wordle, and the occasional Monopoly tournament.
Research code, full-stack products, and tools I actually use. Filter by category below.
Research framework and Python package for computing optimal voting rules under epistemic assumptions. Built positional-scoring optimizers using simulated annealing, gradient descent, and mixed-integer programming over large-scale preference datasets. Co-authored paper accepted at AAMAS 2026.
Async prediction-market arbitrage scanner across Kalshi and Polymarket. Fuzzy event-matching (Jaccard + Levenshtein) hits >95% accuracy without shared event IDs; fee-adjusted EV strategy executes in sub-2s end-to-end.
RAG pipeline over 1000+ FIA regulation and steward-decision PDFs that explains Formula 1 incidents with rule-grounded citations. GenAI Ops experiments on chunking, embeddings, and context size lifted citation coverage from 60% to 93%.
Full-stack web app that classifies fish species from user-uploaded images. RESTful Flask backend integrates Fishial + Groq AI APIs with a persistent identification log; JWT + bcrypt auth, split deployment on Vercel and Render.
Serverless AWS data pipeline ingesting NOAA global Sea Surface Temperature datasets at scale. Glue + Athena ETL workflows enable structured querying and low-latency retrieval for real-time visualization of coral bleaching hotspots.
Browser-based game where players navigate from a random Wikipedia article to a target by clicking only internal links. Bidirectional BFS reveals the shortest possible path after each round. Fully client-side, no backend required.
Web audio player streaming 131 real NOAA hydrophone recordings from eight marine sanctuaries — whale song, snapping shrimp, hurricanes, scuba bubbles, and more. Filter by category and sanctuary, explore via an interactive map, and listen offline via service worker.
F1 telemetry fingerprint tool that visualizes and compares throttle behavior across circuits using 360° resampled data from the OpenF1 API. Side-by-side circuit comparisons reveal performance DNA unique to each track.
AI-powered live-music booking marketplace for New Orleans venues. Multi-agent pipeline orchestrated with BoodleBox, server-sent events for live updates, and Playwright-driven enrichment of venue data.
Research, teaching, and the rooms where I get to actually ship things.
TA for CMPS3160 Intro to Data Science (Spring '26) and CMPS1500 Intro to CS I (Fall '25). Facilitate weekly lab sessions and office hours for 50+ students; collaborate with faculty on course materials, assignment design, and grading rubrics.
Co-authored a paper accepted at AAMAS 2026 on optimal voting rules. Designed positional-scoring optimizers using simulated annealing, gradient descent, and mixed-integer programming over large preference datasets. Contributing to the optimal_voting Python package and Django interface used by researchers across four social welfare functions.
Shipped a production-grade NLP pipeline using custom SpaCy NER to extract structured fields from noisy OCR across financial and identity documents — achieved ~97% precision and automated KYC + form-digitization workflows. Built a multi-format bank-statement processor that emits query-ready Excel reports for downstream data ingestion.
Worked under the Head of Tech Acquisitions at a venture capital fund, researching early-stage tech startups and generative AI applications across industries. Gained exposure to Web3 and blockchain due diligence. Concluded with a comprehensive GenAI landscape report presented to the VC team.
Open to SWE and AI/ML internships. Always happy to talk research, projects, or coffee.