The CTO seat
Technical leadership one to three days a week. Architecture calls, hiring, roadmap, and the decisions your team needs made this week rather than next quarter.
I set the technical direction, then stay and build it.
Fractional CTO for regulated businesses. I find the AI work worth doing, then build it.
Current engagement
A Baltimore wealth management firm, around $2bn under management
Former engagement
A compliance software platform serving the United States defence industrial base
Most technical advice never has to work.
Consultants hand over a strategy deck and leave. Agencies build to the brief and move on. Neither is still there in month nine, when the architecture meets real load and real headcount.
I set the direction and I stay to build it. That means the advice has to survive contact with delivery, because I am the one delivering it.
What I take responsibility for
AI and business change
Most AI work stalls between pilot and production, and rarely because of the model. I pick the few things worth doing, then change the process around them so they stick.
Architecture and platform
Cloud-native systems, designed for the team you actually have.
Team and hiring
Structuring the engineering org, writing the scorecards, sitting in the interviews.
Technical due diligence
Ahead of an acquisition or an investment. What the codebase, the team and the roadmap actually support.
Delivery
Building it in-house, or governing the agency that does. The same accountability either way.
Some of the work
Compliance software for the defence industrial base
AI in an engineering team, without breaking the compliance boundary
Dropping agents into an undefined process just makes mess faster, and in a codebase inside a federal authorisation boundary an unreviewed change is a compliance event. So the order was deliberate: define the delivery process, hire against it, then introduce the tooling. Compliance inspection sits inside the process as a stage gate rather than an audit bolted on afterwards.
Wealth management
Replacing the spreadsheet a whole firm depended on
Decades of client financial planning sat in one Excel model that everybody trusted and nobody had documented. Extracting the assumptions that only existed in the heads of the people who built it took longer than writing the replacement. The result is a working web application, built rather than bought, carrying the firm’s real client models.
Wealth management
AI adoption in a firm that employs no engineers
Thirty people, no technologists, and every client-visible document already passes compliance review. Enthusiasm that quietly runs out of steam is the real risk here, far more likely than a security incident. Policy and training came before tooling, on two plain rules: approved tools only, and you own the output. AI changes the speed of a draft, not the review it goes through.
Three ways this usually works
1 day / week
Advisory
A standing technical seat for leaders making calls they have not had to make before. Architecture review, vendor decisions, hiring input.
Firms with no technology leadership in post
2 to 3 days / week
Embedded
Inside the business. Owning the architecture, running the delivery cadence, and building the technology function while it ships.
Firms replacing systems that carry regulatory weight
Fixed scope
Scoped engagement
A diligence exercise, a platform migration, or governance of an outsourced build. Defined start, defined end, written output.
Diligence, migration, or agency oversight
Rates depend on cadence and commitment. I will give you a number on the first call.
Thirty years, still writing code
I have spent thirty years in technology, most of it inside businesses where a technology failure is a regulatory event as well as a commercial one. Investment banking, global asset management, professional services, defence-sector compliance software.
I still write code. That is unusual for someone who has spent most of that time in senior leadership, and it is rather the point. The gap between a strategy document and a working system is where projects die.
I work as an embedded chief technology officer rather than an external adviser. Engagements are limited in number so that each one has my direct involvement.
Before any of that I shipped a game on the cover of Acorn User, sold a curriculum planning tool at BETT, and wrote my final year project in parallel C and Occam on a Meiko computing surface. It says roughly the same thing about me now as it did then.
- 30
- years in technology
- 22
- of them at one bank
- 10
- acquisitions integrated
- ~300
- applications moved to AWS
Where I have done it
2026
Si Cruse Consulting Ltd
The UK company through which engagements now run.
2025
Embedded CTO engagements
Concurrently: chief technology officer at a wealth management firm, and chief technology officer then technical fellow at a defence-sector compliance software company.
2014 to 2024
In-house technology leadership
Director of architecture at a global asset manager. Chief enterprise architect at a private-equity-backed accounting firm, through ten acquisitions. Co-founder and CTO at Everview, where I built the platform and the iOS app myself.
1995 to 2017
Credit Suisse, twenty two years
From contract programmer on front office trading systems to director of strategic programmes, reporting to the chief architect and the CIO. Core C++ libraries, then the development toolchain used by every developer in the firm.
1989 to 1995
Before the career started properly
Ran a small software company through university, and shipped commercial work alongside it.
Book a call
Tell me what you are dealing with and I will come back with a couple of times that work. No pitch deck, and no obligation to go further.