Through the Lens
Places. Moments.
Away from the keyboard — landscape, street, and travel photography. Italy, Mexico, and wherever the light is interesting.
API Developer · AI Systems · Oakville, ON
I'm an application engineer and API developer at CIBC who has spent the last few years building AI systems to change how development teams actually work — from multi-agent tooling to knowledge pipelines, inside one of Canada's most regulated environments.
Selected Work
A system of four coordinated AI agents that transforms how the engineering team handles daily work. An API Knowledge Extractor converts source code into shared JSON and Markdown manifests — a living knowledge layer the whole team builds on. A PR Review Agent pulls Confluence, Jira, and GitHub context to produce structured approve/flag summaries, cutting routine review time by 50%+. A Release Management Agent monitors the full release lifecycle. An API Modification Agent handles defined change workflows. All built within strict banking compliance constraints — no external model access.
Built locally to validate the approach: embeddings and vector search over a corpus of internal API documentation, enabling natural-language queries over specs that were previously only navigable by reading source. The constraint at CIBC is access to external embedding models — this PoC proves the concept works and defines the architecture for when that gap closes.
A net worth tracker and investment advisory helper that gives AI models persistent financial context — so decisions are made with the full picture across sessions, not just what's shared in a single prompt. Ingests financial statements and converts them into structured metadata tracking net worth, debt, investments, and savings over time.
Designed and built an IoT prototype for Bell's Business Unit IoT Centre of Excellence, with architecture reviewed by Azure Architects. Raspberry Pi 4 with ultrasonic sensors and Electronic Shelf Labels pushing real-time telemetry to Azure IoT Hub — processed through Azure Functions, stored in CosmosDB, and visualised in PowerBI.
Background
I'm an application engineer and API developer at CIBC, where I've spent five years building and maintaining API infrastructure — and the last two years rethinking what that work looks like when you hand parts of it to AI.
My current focus is building AI systems that make the development team more effective: agents that automate the repetitive overhead of API work, knowledge pipelines that surface context instantly, and tooling that reduces the cost of high-quality code review. Working inside a Big-5 bank means doing this with strict governance and no access to external models — which turns out to be a more honest proving ground than most.
Before CIBC I worked across IoT infrastructure at Bell and e-commerce front-end at The Source via CGI — and briefly served as CIO of a startup, which taught me more about real architectural trade-offs than any textbook.
Through the Lens
Away from the keyboard — landscape, street, and travel photography. Italy, Mexico, and wherever the light is interesting.