AI Products & Engineering

NUS-ISS MTech in Artificial Intelligence Systems · Penn State Computer Science

I’m a computer science graduate and NUS-ISS MTech AIS student building AI products across product design and technical implementation. My experience spans AI camera product work, voice-system integration, and AI-assisted prototyping. I’m interested in how user needs, business constraints, and system architecture shape what gets built.

Seeking a full-time Singapore internship · March–August 2027

AI Product · Applied AI · Product Engineering

LinkedIn · Product & engineering experience

Lightmeta · Product Intern, App · May–August 2026
Translated a hardware-first strategy into an instant-capture experience for LumaQ. Led two camera effects and iPhone Camera Control through launch, with PRDs, scope decisions, and cross-functional reviews. Selected decisions →

Penn State · Industry-sponsored capstone · August–December 2025
Led the Azure speech pipeline in a three-person work-instruction assistant project. Integrated a model trained by teammates and supported the move from glasses to a laptop prototype under semester constraints. My contribution →

Earlier: mobile OS issue reproduction at Xiaomi and Python debugging support as a Penn State teaching assistant.

I defined a paper-first maths coach using photo input and one-step prompts, and directed AI-assisted implementation. A key question is how to distinguish finishing with help from attempting a new problem independently; the current adult preview makes those stages visible.

Product decisions & iteration · Run the demo

The brief came from my own Thai delivery decision in Singapore: discover beyond familiar recommendations without losing confidence. I set the problem and requirements; AI agents implemented the prototype. A search result is only a candidate until missing delivery and budget checks are resolved.

Problem & product choices · Run the demo

I defined the requirements and human review workflow for an AI-built release-evidence checker. Deterministic checks inspect declared files and fields; semantic suggestions stay advisory. Its public ThinkBud example shows what the tool found and where the detector falls short.

Inspect the public example · Run the demo

These independent prototypes use AI-assisted implementation. Each repository documents its technical verification and current limits; user outcomes remain to be tested.

• Stock Portfolio — Exact daily portfolio calculations before optional AI interpretation.
• Codex Notch — An unofficial macOS companion for noticing actionable coding-task changes.
• Nianxing · 念行 — Capture a thought, review the draft, and confirm what gets saved.
• FrameText — My independent camera-effects demo, with palm tracking and consistent preview-to-export behavior.

My work spans product requirements, interaction design, scope trade-offs, Azure speech integration, and prototype evaluation. I pay particular attention to data flow, component responsibilities, and failure behavior. My independent web projects use AI coding agents extensively; my capstone contribution includes hands-on voice integration. The linked cases make the division of work explicit.

Mandarin · English | Connect on LinkedIn →