Services
Three ways to work together. You set the pace.
Every engagement is defined by what you hold at the end, not by a calendar I impose. You decide how many sprints to run and how long each one is; I run the work to your delivery standards, with security and governance built into every stage. And when the right answer is software you do not have yet, I build it — purpose-built applications that integrate with the systems you already run — and I stay to support it after go-live.
Not sure which fits? Start with the call — it is the same 30 minutes either way.
How this works in practice
Your cadence, your standards.
If your PMO runs two-week sprints with a Thursday review, that is what we run. If you need a stage-gate plan with steering-committee checkpoints, that is what you get. The program fits your organisation, not the other way round.
Security and risk in the room from day one.
Your CISO, your risk team, and your auditors are stakeholders from the first session, not a gate at the end. Controls are agreed before anything is built.
One accountable owner.
I run the program: scope, schedule, vendors, stakeholders, status. You get one person to call, and a weekly view leadership can read.
Built, connected, and supported.
Where an engagement needs an application — a workflow tool, a customer portal, an AI assistant over your own documents — I build it to work with your identity provider, your data platform, and your line-of-business systems through their APIs. No rip-and-replace. And I do not hand over a repository and leave: I support what I build, fix what breaks, and extend it as your business changes.
01 · Find where AI pays
AI Opportunity Sprint
Who it is for:
Leadership teams that know AI belongs on the roadmap and do not yet know where it creates value, what the data supports, or what security will allow.
The problem it solves:
The roadmap stays a slide because nobody owns the shortlist. Ideas arrive faster than anyone can evaluate them, and the ones that get funded are the loudest, not the best.
What you hold at the end:
- A prioritised shortlist of AI opportunities, ranked by business value, data readiness, and risk.
- A readiness assessment covering data, security, compliance, and operational ownership for each.
- A realistic delivery plan for the top one or two, with the controls your security team will require already named.
- An executive readout your leadership can decide from.
How it runs:
A defined number of sprints that you choose. Each sprint ends with a written output and a review with your stakeholders. Most clients run it as a short, intense cycle; some prefer to spread it across a quarter alongside other work. Both are fine.
Starting at $7,500.
Book a call about the Sprint(opens in a new tab)02 · Build it and ship it
AI Prototype to Production
Who it is for:
Teams with a working prototype that cannot get through security review, governance, or operations — and leaders who have a clear use case, no prototype at all, and need the application built, integrated, and supported from day one.
The problem it solves:
The demo worked. Production means identity and access, data handling, auditability, monitoring, change control, vendor terms, and someone to run it on a Tuesday. Most pilots stall on exactly those things, and most data scientists should not be the ones solving them.
What you hold at the end:
- A production readiness review: what is missing, what is risky, what it will take.
- A purpose-built application — web or mobile — integrated with your identity, data, and line-of-business systems, built to your security standards and owned by you.
- A hardening and governance plan agreed with your CISO and risk team.
- The system in production — access controls, data handling, monitoring, and runbooks in place — delivered under your change-control process.
- A handover to the people who operate it, with training and a support model.
- Support after go-live: fixes, monitoring, and enhancements under a support arrangement we define together, so the system keeps working as your business changes.
How it runs:
Scoped around your pilot and your environment. We agree the sprint structure and checkpoints at the start; I lead delivery through go-live and stay through the first operational period you define.
Starting at $30,000.
Book a call about Prototype to Production(opens in a new tab)03 · Own the program
Fractional AI Lead
Who it is for:
Organisations with more than one AI initiative in flight — or about to be — that need an accountable leader without hiring a full-time executive.
The problem it solves:
AI work is scattered across teams, vendors, and side projects. Nobody owns prioritisation, nobody aligns it with security, and leadership cannot see the whole picture.
What you hold at the end of each month:
- One prioritised portfolio with clear owners, status, and risks.
- A standing governance forum with security, risk, data, and business stakeholders.
- Vendor and platform decisions made with your interests in the room.
- Board-ready reporting on progress, spend, and risk.
- Ongoing support and enhancement of the applications we have built, alongside the rest of the portfolio.
How it runs:
A defined monthly engagement — days per month and the forums I attend are agreed up front and can change as the portfolio does. Minimum term is agreed together, not imposed.
Starting at $8,000 per month.
Book a call about the Fractional AI Lead(opens in a new tab)
A good fit — and an honest "not for you"
Good fit:
- A technology, security, or risk leader with authority to act.
- A real problem or initiative, not curiosity.
- Willingness to give me access to the stakeholders, data, and security people the work depends on.
- Mid-market to enterprise organisations in North America, the UK, or the EU.
Not a fit:
- Staff augmentation or an extra pair of hands inside someone else's plan.
- Projects that require claims about AI that the technology cannot back.
- Engagements where leadership will not engage with security and risk.
- Equity-only or speculative work.
What I build, connect, and support
Everything below is delivered inside one of the three engagements, not sold separately — and supported after it ships.
- Custom applications.
- Internal tools, customer portals, workflow systems, and mobile apps, built for your environment rather than adapted from a template.
- System integration.
- Connections to your identity provider, ERP, CRM, ticketing, calendars, and payment systems through their APIs, with access controls and audit trails in place.
- AI over your own data.
- Assistants and search that answer from your documents and records, with retrieval scoped by role and every answer traceable to its source.
- Workflow automation.
- Multi-step processes — intake, review, approval, notification — run by software with a person accountable at each decision point.
- Data platforms and reporting.
- Cloud data platforms, pipelines, and dashboards leadership actually reads.
- Security hardening and governance.
- Identity and access, secrets, logging, monitoring, change control, and the evidence your auditors will ask for.
- Ongoing support.
- Monitoring, fixes, upgrades, and enhancements for what we have built, on terms agreed up front — so you are never left with software nobody owns.
Questions leaders ask before the call
"We already have a data science team. Why do we need you?"
You probably do not need me to build models. You need someone to get what they build through security, governance, and operations, and to run it as a program. That is a different job, and it is the one I do.
"Will security slow this down?"
It slows things down when it is brought in at the end. Brought in at the start, it removes the late surprises that kill pilots. My background is on the security side; I know what your CISO needs to say yes.
"Do you work with our existing vendors and platforms?"
Yes. I am not selling a platform. I work inside your environment — Azure, AWS, your identity stack, your data platform — and with the vendors you already have or choose.
"Can you build the application, or only advise?"
Both, and building is most of what I do. I have shipped nine products of my own, from scheduling and voice-receptionist tools to a quality-management system for regulated manufacturers, and I have integrated platforms inside banks and healthcare companies. If the work needs software, I build it, connect it to what you already run, and support it afterwards — under your security standards, with your team able to operate it.
"What happens after go-live?"
That is where most AI projects quietly die, so it is written into the engagement. We agree a support arrangement before launch — monitoring, fixes, upgrades, and a path for enhancements — either as a defined support term or as part of the Fractional AI Lead. You get one person who knows the system because he built it.
"How is this different from a big consultancy?"
You get one accountable person who has done this inside banks and regulated companies, not a rotating team. And you set the cadence.
"What happens on the 30-minute call?"
We talk about what is stuck and why. I will tell you whether I can help and what I would propose. If I am not the right person, I will say so.
Start with the call.
Thirty minutes. No deck. Bring the initiative that is stuck — or the application you need built.