Timothy Lau Email & Phone Number

Calgary, Alberta, Canada

Toronto, Ontario, Canada

Skills: Python · Machine Learning · Data Science · Project Design · Research and Development · Behaviorial InterviewsQuadFi is a certified B-corp fintech that aims at utilizing machine learning in providing affordable financial aid and services to less financially able individuals such as students and immigrants.As part of the AI team, I was responsible for the design, development, and implementation of some of the AI code infrastructure and models for QuadFi's products, and… Show more Skills: Python · Machine Learning · Data Science · Project Design · Research and Development · Behaviorial InterviewsQuadFi is a certified B-corp fintech that aims at utilizing machine learning in providing affordable financial aid and services to less financially able individuals such as students and immigrants.As part of the AI team, I was responsible for the design, development, and implementation of some of the AI code infrastructure and models for QuadFi's products, and directing the engineering team on how these were to be implemented into the platform.Projects I was responsible for:Universal Credit- Designed, developed, and directed implementation of a solution to thin Canadian credit data of newcomers to Canada -- a direct improvement to the "Newcomer" product- Developed code base that parsed foreign credit reports, generalized and formatted data for use in financial model- Finalized product as Flask API for engineering team - Directed engineering team for implementation of solution into platformQuadFi Model- Continued development of QuadFi's code base for the overarching rule-based decision making model for loan applicants- Designed and implemented incorporation of foreign data from Universal Credit into decision making process- Supported improvement designs for internal cash flow model Financial Model - Continued refinement of the financial model that I developed as an intern; used in predicting loan default in an applicant utilizing alternative financial data- Responsible for internal and external documentation for both mechanics and performance of the model- Responsible for bringing interns up to speed on code baseAdditionally responsible for:- Collaboration with operations team in decision making process for corner case applicants- Interviewing and participating in decision making process for intern positionTech Stack: - Python, DS libraries- Flask- VSCode- Git, Github- JIRA, Confluence Show less

Toronto, Ontario, Canada

Skills: Python · Machine Learning · Pipeline development · Research and DevelopmentAs an intern, I was responsible for delivering an ever-evolving ML use case pertaining to loan default prediction utilizing alternative financial data. This use case contributes to a higher level model used for processing loan applicants. - Developed and maintained machine learning pipeline for streamlined, rigorous experimentation process, utilizing models such as Logistic Regression (industry… Show more Skills: Python · Machine Learning · Pipeline development · Research and DevelopmentAs an intern, I was responsible for delivering an ever-evolving ML use case pertaining to loan default prediction utilizing alternative financial data. This use case contributes to a higher level model used for processing loan applicants. - Developed and maintained machine learning pipeline for streamlined, rigorous experimentation process, utilizing models such as Logistic Regression (industry standard) and XGBoost- Developed clean, modularized OOP production ready repo for yaml configured ML pipeline- Developed rolling window time series, ensemble method implementations for model training- Experimentations with different model implementations, pushed model going into production- Maintained and organized documentation, model cards, experimentations- Explored model explanability with SHAP- Performed ad-hoc research and initial modeling experimentation into expanding other ML use case scopes (auto loans, education data, +)- Created cash flow static pool analysis -- statistical modeling on simulating loan default, prepayment rates and subsequent losses for different loan portfolios utilizing transition matricesTech Stack:- Python, VSCode, yaml- Excel- Git, GitHub- Google Cloud Platform Show less

Calgary, Alberta, Canada

Delivered ML use case for a large financial institution client - use case allowed better data-driven business decisions and customer micro-management, as well as provided insight into further areas of exploration. Tangible ROI analysis on the use case estimated about $1m per quarter return, with the potential for more if areas of exploration prove fruitful. - Developed and delivered predictive machine learning model pertaining to profitability KPIs - Continued and finalized… Show more Delivered ML use case for a large financial institution client - use case allowed better data-driven business decisions and customer micro-management, as well as provided insight into further areas of exploration. Tangible ROI analysis on the use case estimated about $1m per quarter return, with the potential for more if areas of exploration prove fruitful. - Developed and delivered predictive machine learning model pertaining to profitability KPIs - Continued and finalized development on data acquisition pipeline to utilize data from BigQuery - Developed model pipeline for streamlined experimentation process and production-ready repo - Worked directly with client subject matter experts on feature engineering - Brought previous cohorts' baseline model performance from 30% to 95% - Explored model performance and explanability with SHAPOther deliverables for AltaML: - model cards, experimentation logs, and research on potential time series implementation - "next steps" recommendations for future cohortsOther deliverables for client: - bi-weekly status updates - non-technical reports of problematic features, model impact and residuals for business executives - actionable recommendations - user journey maps, user dashboard - final presentation for stakeholders - return on investment analysis on use caseTech Stack: - Python, Jupyter - SQL - Git, Gitlab - Google Cloud Platform, BigQuery, DataStudio Show less

Edmonton, Canada Area

- Worked on a research project on predicting the condition of a post-surgery ICU patient by modelling data on arterial blood pressure - Collected, organized, and researched papers pertaining to efficacy of different statistical and machine learning models to this domain- Cleaned, manipulated data using R and Python- Worked with various data science modelsTech Stack:- Python, Jupyter - R, RStudio- PostgreSQL, MIMIC III

Bachelor Of Arts - Ba, Economics

Minored in Mathematics Certificate in Economics and Management of Natural Resources, Energy, and the Environment

Diploma, Data Science

3 month bootcamp on Machine Learning, Deep Learning applications and techniques

Diploma, Data Analytics