I worked on the Exchange Connectivity team, building and maintaining low-latency, high-throughput market-data systems in C++ and Java that supported all exchanges IMC trades on.

• Upgraded the NASDAQ PHLX market-data feed from TOPO (BBO-only) to Depth (full-depth order book), implementing new C++ feed handlers and Java processing stages directly from protocol specifications. The new pipeline reduced message-processing latency and improved throughput across downstream consumers.
• Collaborated with a researcher to add a research-critical signal to IMC’s internal market-data protocol. Designed a backward-compatible, extensible schema and implemented C++ and Java components to encode, transport, and consume the new field across multiple latency-sensitive services.
• Led design reviews and coordinated a staged rollout that allowed existing consumers to migrate at their own pace, without breaking production.
• Integrated a new exchange into IMC’s latest connectivity infrastructure using modern C++ template abstractions to factor out shared FIX session-management logic, keeping venue-specific code small, expressive, and zero-overhead at runtime.

For my capstone, I worked with RBC on a system to forecast merchant transaction volume and analyze community profiles using real RBC banking data.

• Built a web-based analytics platform with a Flask backend, processing 1.5+ TB/day of transaction data using Apache Spark, Python, and SQL on Linux-based infrastructure.
• Designed and trained regression models (Decision Tree Regressors) that combined demographics, business metadata, and financial activity to forecast demand for merchants.
• Used agglomerative clustering to group dissemination areas (DAs), enabling more accurate predictions and more interpretable segments for stakeholders.

Science Rendezvous is a national science outreach festival. At Western, I helped build the first dedicated mobile app for the event and later led the technology side of the program.

• Co-developed the first mobile app for a Science Rendezvous event in Canada using the MERN stack (MongoDB, Express.js, React.js, Node.js).
• Shipped features like venue maps, stand discovery, scheduling, and push notifications for 2,000+ attendees, with 300–500 downloads on event day.
• Volunteered over four years, progressing to Program Committee Lead and founding the Technology Committee to formalize and grow the tech side of the event.