Our Work
Proof, not promises.
Real engagements, told straight and anonymised by sector. Where a number is genuinely ours, we quote it. Where it isn't, we won't invent one.
Case studies
Five jobs, five honest before-and-afters.
Different sectors, same pattern: find where the hours leak, build the fix on tools you already pay for, hand it back working.
AI discovery
Wealth management (regulated, MiFID II)
Which AI ideas are worth it, ranked before a euro is spent.
Before
A regulated firm with a long wish-list of 'could AI do this?' and no safe way to tell the quick wins from the compliance headaches.
After
A discovery engagement that scoped every use case and ranked them by return: email-triggered investment summaries, call transcription with meeting notes, and a client-query assistant, sequenced so the safe, high-value ones went first.
A prioritised roadmap the firm could actually defend to compliance, not a pilot that stalls at the first audit question.
- AI strategy
- MiFID II-aware
- ChatGPT
End-to-end Power Platform
Construction & fit-out
From first enquiry to labour on-site, in one system instead of six.
Before
Leads on sticky notes, surveys in one inbox, quotes in Word, the schedule in someone's head. Every handover a chance to drop the ball.
After
One Power Platform + Dataverse pipeline: lead intake → scheduling → property surveys → quotation → labour allocation, with Azure OpenAI drafting quotes, sanity-checking data and answering staff questions in an internal chatbot.
The whole job runs on rails the team can see. No more 'who's doing what Thursday?' by text at 11pm.
- Power Apps
- Dataverse
- Power Automate
- Azure OpenAI
SharePoint + document AI
Legal firms
Case files that behave, and drafting that starts at the second draft.
Before
Matter documents scattered across drives, versioned by filename, and hours lost summarising correspondence by hand.
After
A SharePoint-based case-management system that keeps everything in its place, with ChatGPT summarising documents and email threads and getting first drafts on the page.
Fee-earners spend their time on judgement, not on finding the right version of a letter.
- SharePoint
- ChatGPT
- Document AI
Automation + governance
Public sector (health, ITIL 4)
Change approvals that move faster, with a paper trail that stands up to audit.
Before
Change approvals triaged by hand, post-CAB changes distributed manually, and no clear picture of where the resourcing pressure actually was.
After
Approval-triage automation, a Python tool that auto-distributes post-CAB changes, and a 2019–2026 change-ticket heatmap that shows resourcing at a glance, all wrapped in an AI/automation strategy covering the EU AI Act, GDPR, DPIA, HIQA and a human-in-the-loop framework.
A governance model built for a regulated environment, humans in the loop by design, plus a 2019–2026 change-ticket heatmap that turns 'we're stretched' into something you can actually resource against.
- Python
- ITIL 4 Service Transition
- HITL governance
- Power BI
Curriculum + delivery
Higher education / professional training
The AI course we teach, taught to your team.
Before
Staff hearing 'just use AI' with no idea which tool, which task, or where the line is.
After
Built the AI/ML curriculum for a professional academy and delivers it to live cohorts (ChatGPT Productivity, AI for Business and GenAI), the same practical sessions offered to client teams.
People leave able to open the tool on Monday and use it well, not nodding along to a vendor deck.
- UCD Professional Academy
- Live cohorts
- Copilot certified
Why no logos or euro figures?
Because most of this work sits inside regulated firms and the public sector, and because a saving we can't stand over isn't proof, it's decoration. If a client's happy to be named, we'll ask them, not assume.
Our framework
The same four bridges under every job on this page.
Most AI projects don't fail on the tech. They fall into the gap between a clever demo and a team actually using it. These are the four spans we build across, every time.
- Design & explainabilitymake it legible
- Leadership & trustbring people with you
- Operations & implementationship it into the day job
- Strategy & infrastructurebuild for what's next
- Design & explainabilityAI you can open up and explain, not a black box you have to take on faith.
- Leadership & trustGetting the people who sign it off on board, and the people who use it comfortable.
- Operations & implementationThe unglamorous bit: building it, wiring it in, and making it survive contact with Monday.
- Strategy & infrastructureThe plumbing and the plan underneath, so it still makes sense in two years.