Put AI to work inside the software you already use
Most of my work is AI and cloud development. I help businesses move past experiments to AI features that run reliably in production: document processing, AI assistants over your own data, and automations that remove hours of manual work. The work is secure, measurable and built into your existing systems.
Sound familiar?
- Staff spend hours reading, summarising or re-keying documents and emails
- Customers and staff can't quickly find answers buried in your own documentation
- You've tried ChatGPT pilots, but nothing has reached production
- You're unsure what's safe to send to an AI model and how to control cost
What you get
- Opportunity review: where AI will actually save time or money in your business
- Proof of concept on your real data, with measured accuracy
- Production integration into your existing .NET, Node.js or cloud systems
- Guardrails: data protection, audit logging, cost limits and human review where it matters
- Handover and documentation so your team can run it
Ways to work together
Every engagement starts with a small, paid, fixed-price first step. That's where requirements are gathered and the scope and cost estimate for the rest of the work are written, so you know what you're committing to.
AI opportunity review
from £1,500
Workshop plus a written report ranking opportunities by value and effort.
AI pilot
from £6,000
One use case built on your data and measured against real work.
Production rollout
Quoted per project
Hardening, integration, monitoring and staff handover.
Common questions
Do we need to replace our existing system to use AI?
No. Most of the value comes from adding AI to the systems you already run, through their APIs or database, so staff keep working in the software they know. I build these integrations into existing .NET, Node.js and cloud systems.
Is it safe to send our business data to an AI model?
It depends on the data and the service. The opportunity review sets out what each use case needs to send, personal or sensitive data is kept out of prompts where it isn't needed, and production builds use business AI services that don't train their models on your data. Audit logging and human review are built in where decisions matter.
How much does AI integration cost?
The AI opportunity review starts from £1,500 and a pilot on one use case from £6,000. Production rollout is quoted per project once the pilot has proved the value. Running costs are billed by the AI provider per request: the pilot measures them on real work, so you know the monthly cost before rolling out, and production builds include cost limits.
What if the AI gets it wrong?
The pilot measures accuracy on your own data against real work before anything goes live. Where a mistake would matter, a person reviews the output before it's acted on, and every result is logged so problems can be traced.
Which AI model do you use?
The one that fits the job, chosen on accuracy, cost and where your data is processed. The integration is built so the model can be changed later without rebuilding your system.
Related work
Automated print-on-demand pipeline for a book publisher
Personalised children's books go from web order to a commercial printer with no manual steps: each order generates a print-ready PDF and is submitted to the printer's API automatically.
Read the case study Own SaaS productAccessDrop
Digital-download delivery for Etsy sellers: checks Google Drive, Dropbox and Canva links and produces branded PDFs with QR codes.
Read the case study Own SaaS productStoryPrompt
A year of weekly memoir prompts by email. Answers are collected without a login and typeset into a finished PDF book.
Read the case studyHave a project in mind?
Tell me what you're trying to achieve, and I'll reply as soon as I can to say whether it's a fit and what the fixed-price first step costs.