cv

This page provides a high level summary of my experience. For a more detailed view of my select experiences, please see my resume above.

Basics

Name Nicholas Fung
Label ML Engineer and Data Scientist
Email contact<at>nickfung<dot>ca
Summary ML Engineer and Data Scientist with a MASc from UofT and a BASc from UBC.

Work

  • 9.2021
    -
    12.2023

    Toronto, Canada

    ML Engineer
    Riskfuel Analytics
    Riskfuel uses deep learning to replicate complex derivative pricers written by quants at banks, with the goal of providing accurate approximations at a fraction of the cost for the purpose of real-time trading and P&L calculations.
  • 4.2018
    -
    8.2018

    Hong Kong

    Software Engineer
    Non-Invasive Surgical Innovations
    NISI develops non-invasive surgical robotics. I was responsible for implementing real-time video denoising, as well as researching Computer Vision techniques.
  • 4.2016
    -
    8.2016

    Hong Kong

    Software Engineer Intern
    Cheung Kong Property Holdings Limited
    CKPH is a multinational company that deals primarily with property development.

Certificates

Skills

Data and ML Tools
Python
PyTorch
Tensorflow
NumPy
numba
Pandas
scikit-learn
matplotlib
DevOps
Docker
Kubernetes
GitHub Actions
AKS
On-Prem
Technical Skills
Data Analysis
Data Visualization
Scientific Computing
Quantitative Finance
Deep Learning
Computer Vision
Natural Language Processing
MLOps
CI/CD
Misc. Programming
C/C++
Assembly
Matlab
Julia
Rust
HTML/CSS
Javascript

Education

  • 8.2019
    -
    11.2019

    Toronto, Canada

    MASc
    University of Toronto
    Computer Engineering
  • 8.2015
    -
    5.2019

    Vancouver, Canada

    BASc
    University of British Columbia
    Computer Engineering

Publications

Awards

Projects

  • 7.2020
    -
    12.2021
    Variational Autoencoders: A Hands-Off Approach to Volatility
    A volatility surface is an important tool for pricing and hedging derivatives. It describes the market price of an option on an asset as a function of the option's strike price and maturity. Raw option data is collected and processed, then used to train a neural network which can be used to model observed volatility surfaces as well as generate synthetic surfaces for the purposes of risk management.
    • Quantitative Finance
    • Machine Learning
    • Generative Modelling
  • 07.2018
    -
    08.2018
    Beating the Odds: Horse Racing in Hong Kong
    Horse racing is a lucrative business in Hong Kong, with annual turnover reaching 305 bn HKD (39 bn USD) in 2022. While the outcomes of each race tend to be largely random, several million web pages were scraped to identify key predictors of winning horses. Although various models was fit, the original data was too high dimensional for meaningful predictions. Future work may include improved feature engineering and dimensionality reduction, improved modelling, and collecting more contextual non-tabularized data.
    • Web Scraping
    • Data Analysis
    • Feature Engineering
    • Machine Learning

Volunteer

  • 7.2017
    -
    4.2019
    Treasurer
    nwPlus
    nwPlus organizes several collegiate hackathons, including nwHacks, one of the largest hackathons in the Pacific Northwest, with over 600+ attendees. I was the Logistics Coordinator and later the Treasurer for the university club.