Jan 2023 – Dec 2023 · 11 mos · San Francisco Bay Area

• Invented Toolscript, a new React-like language capable of representing Retool apps in a reviewable and shareable format, powering AI generation and source control
• Built a Retool VS Code extension enabling customers to edit Retool apps in their own local dev environment, expanding developer adoption
• Shipped Enterprise Cloud, a new product SKU leveraging Temporal to allow customers to use on-premise features such as source control on cloud, totaling 1m+ ARR

Jun 2020 – Dec 2022 · 2 yrs 6 mos · Menlo Park, California, United States · Hybrid

• Led a team of 10 engineers to build a Kubernetes production simulation platform powering 5k+ end-to-end tests and personal development environments for 2k+ engineers
• Handled 10X user growth by designing a multi-cluster architecture that leveraged dynamic autoscaling, Cilium virtual IP addresses, and automatic node failover
• Overcame performance limitations within Bazel, Jenkins, and Docker to shorten developer iteration cycles by 80% and save ~80k engineering hours annually

Jan 2019 – Jan 2020 · 1 yr · Seattle, WA

• Developed a Golang microservice that analyzes config changes and summarizes their impact on compute resources, saving >2k developer hours annually
• Built a data pipeline that consumes statistics from the Golang service and generates dynamic AngularJS web reports for project managers across Google to review

Jan 2018 – Jan 2020 · 2 yrs · Greater Toronto Area, Canada

• Contrary is a talent and research-driven investment firm. We back a small number of exceptional people and the companies they build, from seed to scale.

Jan 2018 – Jan 2019 · 1 yr · San Francisco Bay Area

• Designed a Kubernetes service to predictively adjust EC2 Auto Scaling group sizes to match internal demand for build infrastructure, projected to save 1.3 million in the first year
• Expanded Splunk’s supported platforms to Kubernetes by contributing to a Kubernetes operator capable of launching, scaling, and managing distributed Splunk K8s deployments
• Enhanced the open source Splunk Docker image with a feature that moves infrequently searched data to an S3 bucket, reducing storage costs by 7x

Jan 2018 – May 2018 · 4 mos · New York

• Built a convolutional neural network (CNN) that classified Instagram images as ”professional” or ”amateur” with 98% validation accuracy
• Developed a triplet CNN that ranked Instagram images based on stylistic similarity and integrated this model into a "similar users" suggestion service
• Architected and deployed a production server powered by a load-balanced, autoscaling EC2 group that allowed clients to bypass Instagram's per-IP address rate-limiting

Dec 2017 – Dec 2018 · 1 yr · New Haven

• Created a chrome extension at a Yale hackathon to analyze the political bias of news articles using NLP sentiment analysis
• Later accepted into the CITY accelerator at the Yale Tsai Center for Innovative Thinking and presented research findings to members of US Congress, Facebook Policy Specialists

Apr 2017 – Sep 2017 · 5 mos · Greater Toronto Area, Canada

• Took ownership of front-end & back-end features as part of an Agile scrum
• team. Delivered major stories on time to the satisfaction of LOB, such as:
• Implementation of front-end search feature, back-end query logic, and
• Lucene indexing, for a SQL database of over 10 million records
• 50% search time reduction by creating a client-side location-based caching feature that preloads results based on recently searched locations

Sep 2015 – May 2016 · 8 mos · Greater Toronto Area, Canada

• •Prepared and trained advanced science students in preparation for the University of Waterloo Avogadro Contest
• •Designed proprietary course materials and independently mentored students