Software Development Engineer at Census

Developed and owned full-stack features for a SaaS platform serving 100+ businesses, including Notion, The New York Times, and Figma, leveraging Ruby on Rails, TypeScript, Vue, and Kafka to enhance user experience and AI-driven functionalities. Implemented metrics and alerts using Datadog, proactively detecting and resolving 20+ production issues and identifying 10+ performance bottlenecks, leading to improved system reliability and faster incident resolution.

Software Development Engineer at Amazon Web Services

Designed and implemented a robust end-to-end testing framework in Java from scratch, leveraging AWS Lambda, API Gateway, STS, IAM, DynamoDB and Hydraplatform to support and validate multiple identity providers, ensuring a seamless authentication experience for millions of users. Developed a deployment pipeline using TypeScript and AWS CDK, enabling seamless multi-region deployments across environments such as alpha and beta. Created detailed documentation to facilitate adoption and future enhancements.

Undergraduate Research Assistant at University of Waterloo

Spearheaded the development and optimization of a CUDA-enhanced physical simulator using PyTorch and Python for graph neural networks, achieving a 50% reduction in runtime and significantly boosting computational efficiency and research productivity. Utilized Python for data processing and machine learning, developing algorithms that enhanced model training efficiency and data accuracy.

Software Development Engineer at Untether AI

Optimized U-Net and BERT kernels for ASIC architectures using Python and C++, increasing computational efficiency by 35%. Integrated printf functionality into the simulator of the chip using C++, streamlining debugging and enhancing program execution flow tracking.