Michael Wan
Lead Data Scientist · Mastercard · Vancouver
I build AI systems that optimize enterprise operations, from LLM-powered automation to large-scale optimization engines. I also write interactive explainers on AI and deep learning.
- Tech Job Titles: Same Job, Different Price We read 3,060 job postings from 2026 for ten tech titles in the US and Canada. Titles that ask for almost the same skills can still sit 35% to 60% apart in posted base salary.
- Frontier Frugality The best model costs 10x the cheap one, and agentic workflows re-bill your entire context every turn. How to spend frontier tokens only where the intelligence actually lives.
- Jev: A System One Model TypeSafe's Jev answers with typed decisions and probabilities instead of text. What it is, where it sits next to an LLM, and what the first outside tests found.
Experience
Mastercard
-
Lead Data Scientist
Nov 2024 – Present- Engineered Decision Intelligence Pro (DI Pro) fraud detection platform for credit and debit transactions. Optimized the Databricks ML pipeline and infrastructure, accelerating model training by 3x while significantly reducing operational costs.
OOCL
-
Senior Data Scientist
Aug 2021 – Sep 2023- Led Customer Service Automation — LLM + RAG with LangChain, cutting email reply time from 5–10 min to seconds.
- Built EDI AI Developer for GenAI-driven API migration across 20k+ customers — 82% accuracy, 8,000+ human-days saved.
- Developed Shipment Assignment Optimization via CPLEX — $200M annual profit increase on Trans-Pacific and Asia-Europe routes.
- Built Customer Demand & Profit Forecast with hierarchical ML (LightGBM / PyTorch).
- Led 3 concurrent projects with a team of 12 data scientists.
-
Data Scientist
Sep 2018 – Aug 2021- Collaborated with Microsoft Research Asia on empty repositioning — replaced linear programming with reinforcement and supervised learning.
- Built Network Optimization Engine using genetic algorithms for container disruption recovery, improving accuracy from 75% to 87%.
-
Software Engineer (ML)
Apr 2017 – Sep 2018- Built Empty Repositioning Optimization engine with linear programming — reduced $36M in annual equipment costs (4% of $900M).
-
Software Engineer (Full-stack)
Jun 2014 – Apr 2017- Built core ERP components and customer-facing systems on J2EE/Oracle and MEAN stacks, supporting billion-scale shipping operations.
- Delivered route cost calculator and sailing schedule search engine processing 30M+ routes.
Education
Master of Science — Data Science and Business Statistics