Diagnose · Decide · Deliver · Support
I start with what the business needs to change, work out the fix with the people who own the problem, then build it. I run a company myself, so I know what it costs when the technology is wrong.
Where this is different
In my own business I decide what gets built and I build a lot of it, so when something doesn’t move the numbers that’s mine to answer for. The first thing I’ll ask about your project is what it’s meant to change commercially. The stack comes after that.
What I’m for
Most first calls start with a figure that’s moved and no agreed explanation for it.
The answer’s in data you already hold and can’t read: call recordings nobody has time for, free text somebody retypes into a spreadsheet, and until recently reading that lot was a job for a person with a spare month. Most of the work is data engineering, and AI does less of it than you’d expect. The hard part is picking the number that would change what someone does.
Capture happens at the bench or on the call. The decisions get made on a screen somewhere else. I build both, because a dashboard fed by numbers somebody typed in at the end of a shift doesn’t get used.
How we can work together
Same person behind all three. AI runs through each of them.
Build
Bespoke software, SaaS products and internal systems, built and delivered to production.
Partner
A technical co-founder or fractional CTO to own the technology function: what gets built, in what order, and on what architecture.
Consult
Architecture reviews, AI strategy and technical advisory for difficult or high-value decisions.
Applied AI
An OpenAI subscription won’t fix most of what I’m asked about.
Which technique the job needs, and where it should run, matter more than which model you pick.
The right technique for the job
It might be a plain API call, retrieval over your own documents, a model fine-tuned on your history, an MCP server that gives an assistant live business context, or a database query with no AI in it at all. I’ve built each of those, in my own business or for a client, so I can tell you which you need.
Right-sized, and costed
A narrow job doesn’t need the biggest model on the market. At volume the right one runs on hardware I own.
Data that can’t leave
I run speech-to-text and open-weight models on my own hardware because my business handles children’s data. No vendor sees it and nothing trains on it.
Selected work
Three different projects, with a case study on each. The client is described and never named.
Co-Founder & CEO
Scaling a coding education business on platforms, automation and AI systems built in-house.
Recognised2nd of 100, Club Hub UK Top 100 Children’s Activity Providers 2026 · Innovate UK funded · Xero Beautiful Business Fund global winner 2024
AI & automation engineering
Designed and built the AI systems that automate its workflows and join up its business processes.
Technical Co-Founder
A multi-tenant analytics SaaS, built from an idea to production: a parsing pipeline over free-text records, a statistical pricing engine and AI analysis reconciled against the ledger.
How I work
It starts with the commercial problem and ends with software in production. I run every stage myself, and each one gives you something you can use.
The commercial problem, worked out with the people who own it, and written down.
The fix that’s right for the business, whether it’s something new or a repair, agreed before anything is built.
Working, tested software in production, and a handover that doesn’t need me.
I keep the system running and improving for as long as you want it.
About
Thirteen years building software, and most of the last decade running the business it was for.
“We brought Ashley in to find where the business could run better, optimise it and build what it needed. He delivered far more than we asked for. Processing that took a technician most of a morning now takes minutes. The system he built has helped us win pitches with large corporate clients. And I no longer spend every day in the weeds: I run the business.”
Work with me
I reply within one working day to arrange a free 30-minute call.