From Mainframe and Java enterprise systems in the early 2000s to architecting cloud-native, AI-assisted crew automation today. Aviation, healthcare, financial services, e-commerce — different industries, the same instinct for finding out why a system actually broke.
I architect and lead development on JetBlue’s Deadhead Automation (DHA) system — automating crew deadhead assignments across thousands of daily flights using event-driven workflows on Azure Service Bus and Durable Functions. This is the system I write about on LinkedIn — the SOAP integrations, the silent failures, the 2am debugging sessions are all real, drawn from this work.
Led the re-architecture of DHA from Azure Functions to a structured NestJS / TypeScript microservices platform, improving testability, maintainability, and team velocity. Built SkySOLVER / TOPS IROP message processing pipelines for real-time crew eligibility evaluation during irregular operations — under tight SLA constraints, where a wrong decision has real operational consequences.
“Implemented end-to-end observability using OpenTelemetry — distributed tracing, structured logging, and Grafana Cloud dashboards — reducing mean time to diagnosis on production incidents.”
Designed the Azure-native infrastructure underneath all of it: AKS, CosmosDB, Service Bus, Key Vault, ACR, Private Endpoints — treating security and compliance as first-class requirements, not afterthoughts. Mentor and guide a team of 10 engineers, defining architecture decision records, coding standards, and code review processes. Increasingly leverage AI-assisted development — Claude and GitHub Copilot — for architecture reasoning, code review, and documentation, while staying clear-eyed about where it helps and where it confidently misleads.
A short engagement contributing to platform engineering initiatives, in the months immediately following the move to Sutherland/ATMECS — before transitioning to the JetBlue engagement in Jan 2022.
Co-founded and ran a software consultancy delivering mobile and web application development across e-commerce and services sectors. This was the JavaScript decade — staying hands-on across the full stack as the ecosystem itself kept shifting underneath: Node.js, Express, Vue.js, Svelte, React, PostgreSQL, MongoDB.
Conducted JavaScript and Node.js training sessions for developer groups — translating dense language concepts into practical, project-driven learning. That instinct for teaching from first principles is the same one behind the Observability Series and the tutorials going up on this site now.
Owned end-to-end client delivery — scoping, architecture, estimation, development, production support — which meant being the architect, the developer, and the person who got the call when something broke, all at once.
Three and a half years freelancing independently on web development projects for direct clients, after returning from the US consulting engagements and before co-founding JellyGlass — the bridge between working for enterprise clients and building a company of my own.
Five years consulting for enterprise healthcare and financial clients across the US — the era that taught me what “production” actually means when real money and real medical claims are on the line.
At Blue Cross Blue Shield (Jacksonville & Chicago) — claims processing, FSA/HRA adjudication, enrolment and payment processing. At Northern Trust (Chicago) — trust accounting systems. At Fidelity (Jacksonville) — financial services platform work. At Wellpoint (Saint Louis) — health insurance platform development and maintenance.