Tzuta Lin
Sr. Software Development Engineer - Shopping Foundation AI
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolTzuta Lin is a lead research scientist and seasoned software engineer with 12+ years building scalable ML, LLM and big-data systems across Amazon and enterprise products. He has led teams that designed internal LLM pipelines, SDKs, and tooling to accelerate experimentation and inference for shopping and assistant experiences, and now focuses on AI and agentization for enterprise at Salesforce. Tzuta blends deep hands-on expertise in mobile and Android computer-vision apps (creator contributions to widely used projects like LabelImg and multiple dlib-based Android repos) with production-grade backend, data lake, and serving infrastructure. He founded a consumer-facing app, DriveRecorder, which reached 600k+ downloads and was later acquired, demonstrating product instincts alongside research depth. Comfortable from embedded/robotics work at ASUS to cloud-native platforms at Amazon, he pairs research curiosity with pragmatic engineering that ships. Colleagues rely on him to move complex ML workflows from prototyping into reliable, scalable production.
12 years of coding experience
13 years of employment as a software developer
Master's degree Electrical and Electronics Engineering, Master's degree Electrical and Electronics Engineering at National Taiwan University
Chinese, English
Github Skills (38)
opencv10
c-language10
mobile-development10
python10
image-processing10
imagenet10
command-line-interface10
machine-learning10
image-annotation10
android-development10
java10
javas10
android10
dlib10
cprogramming-language10
Programming languages (10)
Github contributions (5)
tzutalin/Android-Object-Detection
Sep 2015 - Sep 2017
:coffee: Fast-RCNN and Scene Recognition using Caffe
Role in this project:
Mobile Developer (Android)
Mobile Developer (Android)Contributions:33 commits, 8 PRs, 32 pushes in 1 year 11 months
Contributions summary:Tzuta primarily contributed to the Android application, focusing on object detection and scene recognition features. Their commits involved adding demo pictures, modifying activity files, and integrating libraries for machine learning. They also addressed build configurations, including adding x86_64 libraries and updating the Gradle build. Furthermore, they updated the model paths and implemented permission requests.
scene-recognitionncnncoffeedetectionrecognition
tzutalin/ImageNet_Utils
Jul 2015 - Mar 2017
:arrow_double_down: Utils to help download images by id, crop bounding box, label images, etc.
Role in this project:Full-stack Developer
Full-stack DeveloperContributions:35 commits, 31 pushes, 1 branch in 1 year 8 months
Contributions summary:Tzuta primarily worked on utility functions for ImageNet data, developing tools for image downloading, bounding box annotation handling, and label creation. They implemented an argument parser for command-line interaction and integrated a preference utility for saving user information. Furthermore, they added functionality for cropping images based on bounding box annotations and generating training/testing datasets.
pythonboxcropboundingdouble
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.