About

I design B2B software systems and run the programs that ship them. On most projects I end up doing both at once. I write the specification and then build the schedule against it, so when I defend a date I know exactly which part of the data model that date is resting on.

I was the solution architect on a new CRM and stayed with it through delivery. Before that I owned a ticketing platform serving several accounts, and a data pipeline for industrial IoT sensors that carried plant-floor telemetry into a store the business could actually query.

On every one of them the hardest call arrived first. You have to decide what the system is really modelling, and then hold that decision steady while the requirements keep shifting around it.

I hold a PMP, which I treat less as project bureaucracy and more as risk management. It gives me the framework to baseline scope before commitment, sequence dependencies where delays would actually hurt the critical path, and structure change control around technical debt rather than optimism.

My work spans manufacturing and healthcare software. In healthcare, that has meant processing raw HL7 feeds into clinical pipelines, building propensity forecasting models for revenue cycle management, and shipping clinical-grade ML screening tools where the cost of a false negative dominates every architectural call.

I also put AI into systems that have to run day to day. Align is a food-logging app I built for myself and still use every day. It keeps reminding me which parts of a problem genuinely need a model and which parts do not.

Every article here works through the reasoning behind one decision. Start with the writing.