I put a self-hosted LLM into production for a US financial client: 10 fields extracted from a single scanned check at 95%+ accuracy, 3-second latency. Deep .NET architecture experience is what made that deliverable rather than a demo.
I'm a Solution Architect specialising in commercial C#/.NET systems, with a long track record of leading teams and owning technical design. Microsoft Certified Azure Solutions Architect Expert. My work sits where legacy systems meet modern cloud: analysing an aging codebase, designing the target architecture, proving it with a working prototype, and negotiating the plan with the people who have to fund it.
Modernization & migration. Azure Cloud Services to Container Apps. .NET Framework to .NET 10. A Java/JSP accounting platform to ASP.NET Core and React. Azure to GCP, containerised and E2E-tested. I've done these end to end — analysis, options with estimates, proof of concept, delivery.
Cloud-native architecture on Azure. AKS, Container Apps, Functions, Logic Apps, APIM, Service Bus, Azure SQL. Microservices, Clean Architecture and DDD, with EF Core and hand-tuned SQL underneath.
Practical AI in engineering. Claude Certified Architect (Anthropic) — a proctored certification covering agentic architecture, MCP integration, prompt engineering and context management. I use Claude Code, Cursor and GitHub Spec Kit daily, with spec-driven development as the working method. Not as an experiment; as how the work gets done.
Delivery across domains. Finance, insurance, healthcare, logistics — check processing, accounting platforms, patient intake with HL7 and EMR integration, RFID inventory at 30M+ items. Led distributed, multinational teams, including hiring, mentoring and code review.
I'm most useful to organisations carrying a valuable legacy system they can't afford to break, who need someone who can architect the path forward and then actually build it.
Open to connecting with architects, engineering leaders and teams working on modernization or serious AI adoption.
