MS Computer Science at UW–Madison. Currently a Data Science Intern at Cloudflare, joining Amazon as an SDE Intern this fall. Previously 2+ years at Oracle building backend services and data infrastructure. I work on LLMs, ML systems, and the engineering underneath them.
Intelligent-Bistro — A mobile ordering app where guests manage a cart through a conversational AI assistant. Claude parses natural-language orders into validated tool_use calls; UI taps and AI calls share one cart-mutation path, so the model's view never diverges from the user's. Expo + React Native, Express + TypeScript, 45 tests.
CoT-Faithfulness — A behavioral and mechanistic audit of chain-of-thought faithfulness across Llama 3.1 8B, Gemma 3 12B, and Claude Opus 4.6. Four CoT perturbations re-injected via assistant prefilling, plus attention analysis. Faithfulness falls as capability rises (F = 0.49 → 0.30 → 0.09).
Certified-Temporal-GNN — Certified robustness for temporal graph neural networks. Replaces the per-window Bonferroni union bound with one shared edge-flip realization under a single Clopper–Pearson bound: sound under Neyman–Pearson, strictly tighter, and up to +41.7 points certified accuracy on real Elliptic Bitcoin data.
SynthShield — Turns sensitive tabular data into synthetic clones with formal privacy guarantees. T5-small fine-tuned with LoRA under DP-SGD (Opacus), reaching ε≈4 at δ=1e-5 with per-column KL under 0.1. FastAPI + Next.js on Cloud Run, built in 24 hours at CheeseHacks 2026.
- UW–Madison — MS in Computer Science (2025–2027)
- IIT Kharagpur — B.Tech (Hons.), Mechanical Engineering; Minor in Mathematics & Computing (2019–2023)
- Data Science Intern — Cloudflare (Jun 2026 – Present)
- Incoming SDE Intern — Amazon (Fall 2026)
- Open Source Developer — UW–Madison Data Science Institute (Feb – Jun 2026)
- Applications Developer — Oracle (Jun 2023 – Aug 2025)
- Email: jmahajan2@wisc.edu
- LinkedIn: https://www.linkedin.com/in/jeevesh28