Pru Yontrarak is a product designer and full stack engineer.

M.S. Computer Science @ Columbia

Peony line illustration

Case studies

Product design

Optimafly: A Multi-City Optimized Flight Planner

  • Shaped multi-city trip planning with flexible date windows and a guided setup flow.
  • Designed scannable itinerary results so travelers can compare legs, cost, and timing at a glance.
Mobile UX

Columbia Daily Spectator App

  • Led UX for the in-app podcast experience: discovery, playback, and small-screen navigation.
  • Iterated on information architecture and UI patterns so content stays readable and tappable in a news context.
Editorial UX

theSHAFT Feature

  • Redesigned dorm detail pages so room data, media, and comparisons are easier to scan.
  • Tightened hierarchy and labeling so students can shortlist housing with less guesswork.

CS projects

Backend · API

Optimafly: A Multi-City Optimized Flight Planner

  • Building a constraint-based API that assembles cost-minimized, multi-city itineraries from flexible trip windows.
Robotics research

Trajectory Optimization Research @ A²R Lab

  • Benchmarking constraint-handling methods for PDDP on robotic control with a Python and CUDA implementation.
ML · Data

Financial Stock Price Forecasting API

  • Combining real-time market data and news sentiment to forecast stock prices behind a single API.
Reinforcement learning

Foosball Reinforcement Learning

  • Training TQC agents in a custom MuJoCo environment to model competitive foosball dynamics with reinforcement learning.
Research

Exploiting Failure in Evolution Research @ Soros Lab, SRI

  • Stress-testing FI-2Pop variants in Lunar Lander to study how preserving failed individuals shapes search and reward hacking.