AI Engineer - Kipt Feb. 2024 - June 2024
Spearheaded the development of a CTPN model to locate text detection and developed a CNN-based OCR model with over 99% accuracy.
Developed LLM to summarize Chinese Classical Literatures.
Who Am I?
Wei-Chung Lu is a graduate student at the University of California, Santa Barbara. Wei-Chung majors in Computer Engineering, particularly in Applied Machine Learning and Software Engineering. He has worked as an Artificial Intelligence and Machine Learning Engineer for two years at Open AI Fab, Merkle, and Kipt. Especially focus on medical AI, marketing recommendation systems, and cloud computing servers.
Wei-Chung is dedicated to implementing AI models. Dealing with integrating cloud-edge computing to apply machine learning or deep learning models efficiently fascinates him hence he enjoys the challenge of bridging researchers to customers and watching his practical contribution lead to a critical impact on people in this world. He is studying tensor and GPU parallel computing to nail the feasibility of inferring models on edge devices. In the future, Wei-Chung is going to build an efficient solution to implement AI everywhere to improve our everyday lives.
My Skills
Python
C/C++
Pytorch
Tensorflow
SQL
Linux
Education
University of California, Santa Barbara
Sept. 2024 - June 2026
Eletrical and Computer Engineering
Cumulative GPA: 3.9
Soochow University
Sept. 2017 - June 2021
Data Science
Cumulative GPA: 3.7
Dean’s List: 2020, 2021
Work Experience
AI Engineer - Open AI Fab Mar. 2022 - Nov. 2023
Independently developed a MobileNetV3-FCOS to identify children's development with over 80% mAP (mean Average Precision), and 25 fps in each test..
Trained a MobileNetV3-SSD model to detect sock top with over 95% mAP and inference less than 0.01 second
Converted model with Core ML and snapdragon SDK for applying to edge or mobile devices.
Machine Learning Engineer - Merkle Aug. 2022 - May 2023
Refined MLOps on GCP to automate customer tagging projects’ model updates to reduce 80% working time.
Enhanced prediction of collaborative filtering to improve sales performance uplift rate by 5% in A/B Testing.
Mentored two interns to optimize program efficiency to reduce RAM usage and data load time by almost 50%.
Research Assistant - Taipei Veterans General Hospital - Neurology Aug. 2022 - May 2023
Assisted Dr.Jong-Ling Fuh in conducting primary source research on the method for analyzing the risk of developing Alzheimer's disease.
Implemented pre-processing and data mining procedures for MRI images and patient biological data.
Industry-Academia Cooperation - Cathay Life Insurance Co., Ltd. Nov. 2020 - Aug. 2021
Led a team to develop a classification model to confirm if the insurance policy returned by fax is signed with 95% accuracy.
Implemented crawler and OCR to analyze the form images from the website for confirming car registration records of the insured with 98% accuracy and less than 15 seconds.
GA solution Engineer Trainee - iProspect - Data Team Oct. 2019 - Sept. 2020
Developed the backend of the GA Hac (B2B tool for GA health check) and established a Flask API to communicate with the frontend.
Analyzed consumer repurchase cycle to reach out to the target consumer for the cosmetic industry leader and exported visualization results to BI tool (looker studio).
Recent Project
All project source codes can be found in my 
GitHub