I build the SaaS and business systems that companies run on.

I build custom software, SaaS products and the business-operations systems companies run on, delivered on AWS. That spans web and mobile applications, automating manual and spreadsheet-based processes, integrating systems through APIs, and building AI agents on top of those integrations. I have done this for over fifteen years, for organisations from startups to public companies.

Yes, in the right places. AI agents work best wired into real systems and integrations, taking on the judgement calls and messy exceptions that used to need a person, rather than replacing whole processes wholesale. The value comes from targeting specific, repeatable, high-friction tasks first and measuring the result, not from automating everything at once.

I start from the process and the pain, not the technology, and favour small, high-impact, low-risk improvements that build momentum. That usually means shipping a focused first version quickly, proving the value, then expanding. The aim is a measurable business result, not software for its own sake.

Yes. I work with organisations of every size, from early-stage startups and SMEs to public companies. Many of my client relationships have run for ten years or more, which says more about how the work holds up than any pitch could.

Agentic AI describes software agents that carry out multi-step tasks and make decisions on your behalf, rather than just answering a single prompt. In a business setting that means an agent plugged into your systems that can follow a process end to end, handle exceptions, and escalate to a person only when it needs to.

Yes - modernising and extending existing systems is as much a part of the work as building from scratch. That might mean moving a legacy application to AWS, replacing the fragile parts, adding integrations, or gradually re-engineering an ageing codebase instead of risking a big-bang rewrite. Having spent years maintaining production systems, I favour low-risk, incremental change that keeps the business running throughout.

It depends on scope, but you rarely need to build the whole thing at once. I favour a focused first version that proves the value quickly and keeps the initial outlay small, then expanding as it earns its place. Rather than a fixed, all-or-nothing quote, I would scope the highest-impact slice first and build from there. A short conversation about what you are trying to achieve is the quickest route to a real figure.

Security and data protection are designed in from the start, not bolted on at the end. I build on AWS with least-privilege access, encryption and careful data handling as defaults, and I have a long-standing focus on application security. My degree was in human error and human-machine interaction, so I pay particular attention to the mistakes people and systems actually make, and design to be resilient to them.