Ph.D. student in Computational Media at UCSC with a strong foundation in Comp Sci, AI, Data Science, and Game Dev.
As the CTO & Co-Founder of 烁SHUO, I bring hands-on experience in leading innovative tech projects.
Drawing from an interdisciplinary background in Game Design, Comp Sci, AI, and DS, I am eager to bridge these domains to drive innovation and foster collaborations with industry peers.
Courses taken: Machine Learning, Algorithms, Operating System, Distributed System, Computer Network, Computer Architecture, ...
Courses taken: Artificial Intelligence, Algorithms, Database Management Systems, Operating System, Building User Interfaces Project Management, Computer Graphic, Cryptography, Machine Organization and Programming, Computer Engineering, ...
- Design a visual understanding model based on CNN and develop a
task-independent
Meta Learning algorithm that automatically learns neural architectures for
Few-Shot
Neural Architecture Search (NAS)and weights of new tasks.
- Research on Meta Learning and NAS, run the designed model on Few-Shot Image
Classification benchmarks such as Omniglot and miniImageNet, gather data, and
write
research paper.
- Improve the algorithm and paper based on feedbacks from weekly discussions
with
the advisor.
- Designed and implemented a trajectory prediction Machine Learning model with a
team based on Lyft's extensive dataset and l5kit.
- Researched on trajectory prediction models such as CoverNet, MultiPath, and
Multi-agent trajectory forecasting technique to improve our algorithm’s
prediction
accuracy.
- Tested the designed algorithm in CARLA on simulated autonomous driving
vehicles
under near-real-life street conditions for trajectory predictions.
- Wrote weekly research reports and demonstrated prediction results for a
research
group over 30 people.
- Competed in the Lyft Motion Prediction for Autonomous Vehicles Kaggle
competition
with the team.
- Design two web applications for the ESG (Environment, Social, and Governance)
team.
The first is a web application
based on a flexible React.js framework that enables users to retrieve the live
ESG
ratings and rankings for their own
companies from three major ESG rating agencies databases and simulate their rate
by
customizing weights of a classic
ESG rating model.
- The other application retrieves data from the three major ESG rating agencies’
databases, enabling users to search and
compare trends of the various companies’ average stock price within a particular
industry or at a certain rating level in
the past five years.
- Assist the AQI (air quality index) team's project and develop a questionnaire
application based on WeChat applet,
enabling over 400 research subjects in four groups (three experimental groups
and
one control group) to answer questions
simultaneously. These questions are randomly select questions from a database
based
on answerers’ group.
- Design and implement a frugal and user-friendly full-stack web application
with
React, Bootstrap, Spring Boot, MongoDB, GitLab CI/CD pipeline, and AWS, which
helps
current and future Costa Rica refugees to learn migratory regularizations and
rights, as well to prepare for naturalization exams.
- Improve user experience and implement new features based on feedback and
requirements from Costa Rica NGO and government during weekly meetings.
- Designed a safe driving detector with team, which analyzes real-time streaming
videos and detects if the driver is
smoking, using a cellphone, or driving without a seat belt based on face
detection
and gesture analysis techniques.
- Developed the safe driving detector using Python based on OpenCV, TensorFlow,
and
NumPy, and trained by a dataset
with over 102k images and videos retrieved from related websites using Python
scripts.
- Designed and built a human-computer interaction (HCI) website, with HTML5 and
JavaScript, using tomcat and apache,
to help users create, retrieve, update, delete (CRUD) test case data, in order
to
train an AI customer service system.
- Designed and implemented API layer interface connecting back-end and
front-end.
- Improved front-end interface and back-end functions according to users’
feedback.
- Developed and edited SDK with C and Java for Azure products including Azure
IoT
and
Azure IoT Edge.
- Learned and presented Azure Stack and Azure Cloud Technology to other
departments.
- Tested algorithm modules for machine learning of facial recognition and
detectors
for mechanical components.
- Developed Augmented Reality (AR) and Mixed Reality (MR) Industrial Application
with
JavaScript based on Microsoft
HoloLens and PTC ThingWorx Studio platform for the McKinsey&Company's new
program on
Augmented Reality on
the Internet of thing.
- Connected with PTC ThingWorx Studio and update and demonstrated AR and MR
product
on production lines in real
factories with IoT technology.
- Connected with foreign clients from countries such as United States, India,
and
Luxembourg.
- Sought potential cooperative companies and private banks.
- Promoted FinTech products, services, and resources for companies planning to
shift
their business online.
- Manage a team of 20 to design and develop a 3D Metaverse social infrastructure
using Web Assembly and WebGPU engine in BABYLON.js as frontend, Agora as live
stream
feature, WebSocket and Colyseus as backend, MongoDB and Tencent Cloud as
database
and server.
- Research and implement web applications with cutting-edge technology of WebGPU
to
maximize 3D graphic performance on mobile platforms.
- Designed and implemented conversational agent that assist users with their
shopping experience
through connecting DialogFlow with webhook that connects to the web API.
- Designed and implemented a mobile application that allows users to achieve
their
fitness goals using React Native.
- Designed UI and UX through paper, static, and interactive prototyping using
Adobe
XD.
- Implemented accessible features that enables ease of use for people with
visual
impairment.
- Designed and implemented a course enrollment web application with an effective
course recommendation system based on React.js.
- Designed UI components through examining designing principles.
- Enhanced user experience and usability through heuristic evaluation.
- Designed and implemented a data-visualization and image annotation web
application
based on React framework based on React.js and MongoDB.
- Enabled users to present large amount of data in various kinds of charts and
graphs, also allowing users to annotate image of their
choice with polygon and text attributes, to feed their Machine Learning projects
as
image training data.
- Designed and implemented a 2-layer, feed-forward neural network and trains it
by
using the back-propagation algorithm with the ReLU and Softmax activation
functions
to read users' handwriting.
- Designed and implemented a GUI to help users plan their meal based on the
different nutrient information of the various
type of food on the food list on the UI.
- Enabled users to examine food nutrient, total meal nutrient, and add
customized
food to database.
Bay Area, CA