In summer 2026 I interned at Aspire Digital Technology in Guangzhou. Most of my time went to leading the Rokid AR guide project. I also tested AI agents on search, tool use, and scheduled tasks, and wrote up the failures I could reproduce so the team could decide what to rely on.
In September I led a team at the Battle of Schools Hackathon, where we built AllAbout Campus. On my own, I've built a memory-review prototype for agents and a playlist sequencing app called FlowList.
Before university, I spent summer 2024 in a Johns Hopkins summer program analyzing obesity risk factors with Python and XGBoost. That fall I volunteered with a rural education program in Xinshao, China, introducing students to computers, classification, and neural networks. One classroom had no computer for the slides I'd brought, so I explained the basic idea of machine learning with a Peppa Pig robot story instead. I still think about that when I explain technical work: start from what the listener has in front of them.
Tools I use most
Python, TypeScript, Kotlin, and Java; React, Next.js, and Fastify; Android, Playwright, and the Steel SDK.
Contact
jesse.zeng.work@gmail.com · Résumé (PDF) · GitHub · LinkedIn