Applied AI is the discipline of translating AI capabilities into practical solutions that improve how organizations operate, make decisions, and create value.
This lab explores that discipline through practical case studies, engineering projects, and reflections on the architectural decisions behind them.
Every project here begins with a business problem.
Applied AI is more than building intelligent systems. It requires understanding the business, exercising sound judgment about where AI belongs, and designing systems that can deliver value reliably. This lab sits at the intersection of all three.
Every meaningful AI initiative starts here. Understanding the business context, operational challenges, and desired outcomes before considering any technology.
Evaluating whether AI is the right approach, selecting appropriate techniques, and applying them where they create genuine value.
Designing architectures that are reliable, scalable, and maintainable, drawing on years of experience building and supporting enterprise technology.
The case studies rehearse the discipline. Oxaron is where I am practising it on a problem I chose myself, for people who will use it.
Most organisations cannot say what their software actually costs them. Accounts nobody has opened in months stay on the bill, licences sit unused, and tools quietly overlap. Oxaron connects read-only to Microsoft 365 or Google Workspace, prices the whole estate, and lets a team test a new tool against the one they already run before anyone signs for it.
The engineering question underneath it is trust, the same one the case studies keep circling: a figure is only useful if the person acting on it can check how it was reached.
Each case study begins with a real business challenge and works through the reasoning behind the solution, from understanding the business context and evaluating the role of AI to designing the architecture, implementing the solution, and defining how success should be measured.
Analysts at a research firm were spending half of every week reading SEC filings before they could produce any original analysis. The study works through whether AI was the right call, the architecture that makes its answers trustworthy, and how the value would be measured, resulting in a product you can try.
Less like a traditional portfolio and more like deliberate practice in Applied AI engineering. Documenting not only what I build, but how I think.
Each case study takes a business challenge, real or fictional, and works it from problem definition to implementation. It combines business analysis, AI opportunity evaluation, systems design, engineering decisions, and measurable business outcomes.
Browse case studies →Thoughts, frameworks, and questions that shape my understanding of Applied AI, from why some AI initiatives succeed while others fail to how organizations can adopt AI more intentionally.
Read perspectives →Working implementations that bring ideas into practice. Each project explores technical concepts, architectural decisions, and engineering trade-offs that support the case studies.
See the projects →