Sep 2023 – Dec 2023 · 3 mos

• Developed a new feature in our web app using Python (Flask), GraphQL, and React to display a breakdown of users’ tax lots for their crypto assets, helping users and their accountants optimize their capital gains taxes.
• Built functionality to queue CSV download jobs, poll for job completion, and display informative notifications, allowing users to seamlessly download CSVs of their tax lots.
• Shipped tax lots feature to production for internal testing and feedback, with plans to offer it in our highest-paying subscription containing 10% of our user base.
• Wrote robust unit tests using Pytest and the React Testing Library.
• Updated JavaScript fiat formatter to reduce the ambiguity of currency symbols across our app.

May 2023 – Aug 2023 · 3 mos · Redmond, Washington, United States

• Developed Python-based machine learning agents that use reinforced learning to dynamically repair unhealthy nodes in Azure.
• Created a contextual bandit agent with Vowpal Wabbit that leverages diagnostic information and historical observations to learn a repair policy more informed than previous strategies that could reduce VM reboots by 16.1% and VM downtime by 25.8%.
• Considered the interpretability of the contextual bandit agent via feature and feature-interaction weights.
• Merged two multi-armed bandit models that optimize individually for VM reboots and VM downtime into a single hybrid model that automatically balances and trades off these metrics, successfully deploying this model to address customer needs.
• Improved extensibility of existing models through refactoring, while increasing test coverage for expected learning functionality and edge cases.

May 2022 – Aug 2022 · 3 mos · California, United States

• Researched, designed, and prototyped an extensible OS build metadata service in Python for CI systems, with parallelization for production builds resulting in a potential 30 mins - 2 hrs improvement.
• Migrated tooling from outdated infrastructure to the cloud.
• Used FastAPI to develop endpoints that authenticate to, query, and receive events from upstream services; deployed in Kubernetes with Docker containerization; and managed secrets (JWT auth tokens, passwords, etc.) with secure infrastructure.
• Re-implemented a tool in this service based on existing build checks to detect ambiguous file paths during OS image assembly.

Sep 2021 – Dec 2021 · 3 mos · California, United States

• Built user interfaces and their backends with Java, JavaScript, SQL, HTML/CSS, Perforce, Spring, and Lightning UI, improving integration of customer data sources with our app and adding support for custom actions in our app's object manager.
• Contributed to a new self-service app page to offer users a centralized view of asynchronous jobs and the ability to cancel them, positively impacting 14,700+ organizations that create an average of 975+ jobs/day with this feature's release in production.
• Spearheaded development of app page for users to manage JSON that differentiates objects in Amazon DynamoDB tables.

May 2021 – Aug 2021 · 3 mos · Waterloo, Ontario, Canada

• Used Python to expand tool that mines social media data and processes it with machine learning to support qualitative researchers.
• Implemented an NMF machine learning model using Scikit-Learn, improving topic model generation time by 6000+%.
• Developed support for integrating Twitter data with Tweepy: handled authentication, query generation, data retrieval UI/backend.

Jan 2021 – Apr 2021 · 3 mos · Toronto, Ontario, Canada

• Reinforced automated web scraping bots that fetch financial documents for accountants and small/medium businesses, using asynchronous JavaScript, jQuery, Puppeteer.js, Nightmare.js, and Electron.
• Wrote code deployed in Hubdoc's core product with a direct and immediate positive impact on 8,650+ customers.
• Moved scraping connections for 1,130+ customers out of a failing state, decreasing failing users by 75% for a major vendor.
• Expanded and refactored a key web scraper for RBC, fetching 125,000+ new financial documents for 6,670+ customers.
• Tackled a serious issue involving a third-party API from the Bank of America that affected 1,260+ customers.
• Developed and tested a feature for our internal platform (using Backbone.js, Mocha, and Chai) to categorize notes on our roster of 600+ bots, allowing for more effective communication between our software developers and Customer Experience team.

Sep 2020 – Dec 2020 · 3 mos

• Worked in a small team of 3 to build MomsInMind, a web application to reduce postpartum depression in mothers.
• Developed several responsive and reusable React.js components (ex. mobile navigation, video carousel, footer), tested HTTP requests in Postman, and used HTTP requests to successfully connect videos and user profile information to/from our backend.

Jun 2020 – Aug 2020 · 2 mos

• Researched and resolved 22+ customer issues regarding the QNX embedded operating system.
• Reproduced and debugged errors in customer QNX C/C++ projects by working efficiently in a command line and Eclipse environment on a Windows host computer and QNX/Linux virtual machines in VMware Workstation Pro.
• Tested and reported critical errors in Docker setup documentation prior to the major release of the QNX SDP 7.1.

May 2020 – Jun 2020 · 1 mo

• Developed a full-stack Chrome Extension for Twitter using React.js and a Firebase realtime database.
• Users can log in through Twitter and save tweets directly on twitter.com, which can then be sorted by user-defined tags.

Feb 2020 – Apr 2020 · 2 mos

• Developed website components with React.

Sep 2019 – Feb 2020 · 5 mos

• Worked to research, gather and interpret data from (as well as document) thermocouple and retroreflective sensors.

Sep 2017 – Jun 2019 · 1 yr 9 mos

• My role involved managing and calculating event costs, revenues, and reimbursements. I also worked with the executive teams of Visual Arts Council, Music Council, and Performing Arts Council to develop and host Arts Festivals, Arts Weeks, Arts Banquets, and fundraisers. By regularly communicating with our supervisors, I helped ensure effective communication and transparency. Within the council, I led a group of 6-7 members to perform promotional efforts, organize event logistics, and volunteer at events.

Sep 2016 – Jun 2019 · 2 yrs 9 mos

• I managed our team’s outreach sub-team by coordinating sponsorships, engaging the community in STEM workshops, and writing applications for awards. As a presenter for the Chairman’s Award (the most prestigious FRC award) for 4 years, I used my experience to train junior members to present for the Chairman’s Award. In 2017, my efforts significantly contributed to our team winning the district Engineering and Inspiration Award, which allowed us to advance to the Ontario Championships.