[ click something in the room ]
Reliability, usability, and efficiency are the qualities I care about most in any solution I build.
How I work - I’m good at turning ambiguous concepts into clear system diagrams, then building them into products that help people solve problems.
What I care about - The result has to hold up in real use.
Part-time.
With Internship.
A few frames I'm glad I kept.
new york 00
new york 01
new york 02
temple 00
temple 01
temple 02
drugstore 00
drugstore 01
drugstore 02
waterbank 00
waterbank 01
waterbank 02
firework 00
firework 01
firework 02
light 00
light 01
light 02
macau 00
macau 01
macau 02
A few things taking up good space in my life lately.
I like slowing down to catch the small moments — check out my work by clicking the camera.
Food is one of the ways I stay close to people. I grew up with Chinese cooking, so that is my foundation, but I also like borrowing carefully from Japanese fish and beef techniques, Korean appetizer-making, French steak and lamb chops, and Italian pasta and sandwiches.
I also carry the Chinese tradition of 家宴: a home banquet for the family and friends closest to you. To me, it is one of the highest forms of hospitality, because the care is in the planning, the cooking, and the time everyone spends at the table.
I spent a couple of years (2015-2017) volunteering with Edmonton's film and cultural festivals, including the Edmonton International Film Festival and Edmonton Italian Culture Festival. I liked being close to the energy of people gathering around stories, food, language, and place.
(psst — click the mug, or the "Welcome .." 5×)
I implemented a sensor-to-cloud stack for body-temperature monitoring (BLE, Raspberry Pi, MQTT, MongoDB) using React/Node and Google Firestore, and added alert thresholds and a booking flow to close the loop from detection to care.
I reproduced a published indoor navigation method and benchmarked multiple awareness policies (visibility, moving speed), including combined policies, using ROS and C++.
I spent this internship in the Reinforcement Learning & AI Lab, exploring async reinforcement learning workflows for Franka robot experiments: franka-async-rl↗.
At Autonomous Robotic Vehicle Project↗, I worked on software for an underwater robot: the simulator everyone used for testing, object detection with histograms and a random forest classifier, camera calibration, Qt GUI work, and mission planning.
As team lead, I helped 20+ computer engineering/science students work together through GitHub, coding conventions, development process, and weekly scrum. I also taught GitHub workflows to 30+ engineering/science students and presented robotics content at the ROS Edmonton Meetup.
The team ranked 9th of 44 teams at RoboSub 2017↗ and later received 4th of 53 teams at RoboSub 2019↗.
I designed and built ORION, an overnight remediation system that automated data-integrity checks for next-day investment operations, replacing hours of manual work. At its core was a multi-threaded Python scheduler running 3000+ daily jobs, backed by a metadata database for later data-science analysis.
First job! I built Payment Assistant, an internal web app that helped automate compensation filing workflows and supported 100+ weekly uses. The work took me across the React/Redux interface, Material UI components, ASP.NET APIs, and SQL queries.