About
I build AI systems that people can trust — on-device models, agents with human approval gates, and LLM features that fail safely.
I'm a computer engineering student at UCF who builds AI systems end to end — from training the model to shipping the app it runs in. This summer I was an AI/ML engineer at Intrastack Solutions, an AWS consulting firm, building production web systems and AI features for enterprise clients.
On my own time I've trained a Transformer that reads American Sign Language from a webcam, built a local-first agent that operates my computer from my phone, and shipped a full-stack travel platform with streaming LLM itineraries. The thread through all of it: AI that's honest about its limits and asks before it acts.
README.mdarjun.ts
~/arjun/README.md
1/**
2 * Arjun Sherugar — Software Engineer, AI/ML
3 * Now: B.S. Computer Engineering @ UCF
4 * Summer 2026: AI & ML Engineer @ Intrastack Solutions
5 * Interests: on-device models, agents, LLM features that fail safely
6 * Rule: the model proposes, a human approves
7 * Work authorization: U.S. citizen
8 */
Markdown · UTF-8 · LF40F