USC Viterbi · CS & Business · Class of 2026

About Me

I am a graduate of USC Viterbi with a BS in Computer Science and Business Administration. At AMD, I worked on validating the benchmark data sales teams rely on; at Iternal Technologies, I turned AI document tools into demos that helped close customer deals.

Experience

AMD
Software Engineer Intern
May 2025 — Aug 2025 · Santa Clara, CA
Overview

Here I joined the US Cloud Team, which focused on providing accurate benchmark statistics of AMD cloud CPU instances across major cloud providers (AWS, OCI, GCP, etc.). My role began as general support for the team, mainly automating manual processes for uploading benchmark statistics. But with more exposure to the benchmark database, it became clear the real bottleneck wasn't the data itself but the lack of a system to validate it. That shifted my focus from one-off requests to building a process that fixed the underlying process.


Highlights
  • Audited a 28K+ entry benchmark database, found systemic inconsistencies, and proposed a validation system now used by both engineering and sales to evaluate product performance.
  • Worked directly with field application engineers and sales to understand what they actually needed from benchmarking data, not just what was easy to build.
  • Built Python automation that cut manual data processing by 85%, making the dashboards leadership relies on more trustworthy.
cleaner data → sales trusts the numbers they're quoting
Iternal Technologies
Product Management Intern
Jun 2024 — Aug 2024 · Austin, TX
Overview

Iternal Technologies develops AI tools that convert unstructured data into searchable knowledge bases and deploy AI assistants capable of running locally, offline. I partnered with business and engineering teams to define use cases and drive product-market fit for Autoreports, an AI tool for extracting precise insights from large, complex documents. Working at a fast-moving AI startup taught me how quickly priorities shift in pursuit of product-market fit and how to translate ambiguous user needs into shippable features under that pressure.


Highlights
  • Translated raw user pain points in medical and financial document review into prototyped AI features.
  • Produced 15 product demo videos that lifted user retention 27% and customer acquisition 12%.
  • Scoped LLM-powered automation for tax and research workflows, cutting manual review time by up to 4 hours per task.
demo video → 12% more customers said yes

Projects

FEB — APR 2026
WeChat Blog Publishing Bot
Node.js · TypeScript · webhooks · LLM routing
Overview

The Los Angeles Zhongshan Society, a community organization connecting LA residents from Zhongshan, Guangdong Province, ran on a website that only a few technical volunteers could update. This left most non-technical members with no way to post or update community blogs. I led a team of five engineers to solve this by meeting members where they already were: WeCom, the chat app the community used daily. The interesting challenge was turning free-form, conversational chat input into structured, publish-ready content without making it feel like using a CMS.


Highlights
  • Built a message-routing pipeline that interprets free-form WeChat input and tiers it to the right action, cutting AI model calls by 85%.
  • Shipped a bilingual English/Chinese translation layer inside the publishing workflow, cutting content creation time by 63%.
  • Added AI captioning that writes publish-ready image descriptions and auto-inserts them into drafts.
View demo ↗
FOUNDER · 2022 — PRESENT
Codeworks
curriculum design · team leadership · ops
Overview

I led a team of 5 instructors to partner with Pleasanton Unified School District, starting an afterschool program that taught coding fundamentals to underprivileged elementary and middle school students. Being so close to Silicon Valley, there was no shortage of paid coding programs, but those opportunities were limited to families who could afford them. We started Codeworks to close that gap and give students from all backgrounds a place to begin their coding journey.


Highlights
  • Co-founded and designed a 6-week coding curriculum for elementary and middle schooler students, rewriting it weekly based on direct student feedback.
  • Led 5 instructors through weekly syncs to prioritize curriculum changes, driving a 21% lift in student test scores.
  • Owned the full lifecycle for 30 students — recruiting families to enroll, delivering the weekly curriculum, and measuring outcomes to guide the next cohort.

Why Technical Sales

Most of my work has lived in the gap between what something does and what it's worth to the person paying for it. At AMD that meant sitting with engineers on a benchmark database, then building a validation process that strengthened how sales could stand behind the numbers they quoted customers. At Iternal it meant turning a research prototype into a 90-second demo that actually moved acquisition.

I want to do that on purpose, full time, for AI infrastructure: a space where the buyer is usually an engineer, the seller usually isn't one, and the gap between them is exactly where I've spent every internship so far.