Frank Lu Email & Phone Number

Richmond, British Columbia, Canada

- Develop an MLOps workflow that monitors model drifting, trains an ML calibration layer, and packages & deploys the ML layer using MLflow- Build a chatbot backed by Azure OPenAI for quick Q&A on internal documents- Construct a machine learning pipeline that predicts departure delay with 85% accuracy

Vancouver, British Columbia, Canada

- Designed and deployed a machine learning pipeline as REST API that detects reportable cancer reports with 98% accuracy- Developed a Machine Learning pipeline that converts manual medical report classification into an automatic workflow using BERT NLP and RandomForest- Developed an orchestrated MLOps pipeline that automates data transformation, distributed hyperparameter tuning, model staging, and endpoint deployment

Vancouver, British Columbia, Canada

- Develop a Machine Learning pipeline that converts manual medical report classification into an automatic workflow using BERT NLP and RandomForest- Develop a cancer relapse predictor that ingests temporal medical event dataset and identifies high-relapsing-risk patients at 90% accuracy

Vancouver, British Columbia, Canada

- Built end-to-end data pipeline and interactive web app that has been adopted by the client for streamlining 911 response procedures- Reduced redundant SOP files across police agencies using document clustering and topic modelling with NLP and Machine Learning- Improved document retrieval time by 80% by implementing a keyword search function with Gensim library and TF-IDF- Created concise and intuitive visualizations that convert emergency responder instructions into procedural maps

Bellevue, Washington, United States

- Designed and developed a metrics logging and visualizing service for the 2M+ transactions per second Global Pricing Engine- Formulated the data synchronization scheme for cloud-based and on-premise in-memory database clusters and reduced system error by 50%- Constructed a testing framework to detect errors in fundamental business logic in the early stage of product development

Seattle, Washington, United States

- Designed and built a cloud-based data pipeline that aggregates and synchronizes datacentre allocation strategies for 12 time zones- Implemented a signalling platform that pinpoints the error location along the data pipeline and improved failure response time by 20%- Created a data collection service that speeds up the retrieval of customer orders and vendor schedules by 50%- Collected and unified the requirements of customers and stakeholders and defined an API contract to be used across 4 teams

Richmond, Bc

- Integrated SCADA-over-TCP/IP with backup power suppliers to provide uninterruptible power to critical systems in airport- Coordinated YVR, BC Hydro and City of Richmond to upgrade the power and control systems for street lights- Held weekly meetings to track project progress, review change orders, and mediate between stakeholders to ensure a safe, timely and cost-effective delivery of projects

Surrey, Bc

- Prepared a report for BC Hydro that analyzed the performance of various Voice-Over-Satellite solutions used in power grid automation- Set up wired and wireless VLANs with Cisco routers for testing and evaluating Smart Grid systems

Master Of Data Science, Data Science, 88%

Bachelor Of Applied Science (B.A.Sc.), Electrical Engineering, 88%

Graduation with Distinction.

Technologist Diploma, Electrical And Computer Engineering Technology, 92%