Experience & selected work

From building data systems for global companies, to fifteen years running NHS services and budgets, to building AI systems that small businesses use every day.

Selected current work

Oberflacht

Finding high-end shops for a candle maker

Problem

Oberflacht makes handcrafted sculptural candles sold through high-end design and fashion shops, and wants to grow abroad.

Finding new shops that would suit the brand meant researching every one by hand: what they stock, who else they carry, and whether the fit is real.

What I built

A research system that finds shops that already stock brands like theirs, checks each one, and returns a ranked list with the reasons.

The list keeps itself up to date, and an interactive map shows which brands sit in which shops, so new leads come from brands they already sit alongside.

Result

More high-end shops now stock Oberflacht’s candles.

Found by Markus

Sourcing vintage furniture for a dealer

Problem

A vintage furniture dealer whose business depends on finding the right pieces before anyone else does.

Good pieces appear and disappear fast across a huge market, and only the dealer can tell a good one from a weak one. Looking took a large part of his time.

What I built

A system that watches a huge, fast-moving market, applies the dealer’s own criteria automatically, and delivers a shortlist worth his time.

Weak matches are removed before he sees them, so his attention goes only to pieces that are likely to sell. The criteria are his, written down and applied every time.

Result

He now reviews a shortlist instead of the whole market.

Independent publishing site

Running a large publishing site with very little manual work

Problem

A publishing site I built and run, with several thousand pages in two languages.

That many pages need writing, measuring and keeping up to date, far too much to manage by hand.

What I built

Tools that write and manage the pages, pull in Search Console, Analytics and advertising data, and test changes against real traffic before rolling them out.

Then an AI management layer that runs most of the day-to-day content and performance work on its own.

Result

Several thousand pages, around 6,000–7,000 visits a day, running largely without me.

Advertising revenue at an annualised rate of around £10,000.

Background

I started in IT consultancy, building data and reporting systems and prototypes for companies such as Deutsche Bank, Virgin Atlantic and BP.

I then spent fifteen years in the NHS: around five years in project and change management at St George’s Hospital in London, then over ten years at NHS City and Hackney, where I was responsible for children’s mental-health services within a wider budget of around £18m that also covered adult services.

I now work independently from London, building AI systems for small businesses.

Fifteen years in the NHS

In the NHS my role was to work out what a local population needed, decide how the budget should be spent, and design the services to meet that need.

That meant buying and managing services from hospitals, councils, charities and private companies, within tight financial and regulatory limits.

In City and Hackney I led a redesign of children’s mental-health services that improved productivity, value for money, patient outcomes and patient satisfaction.

  • £18mNHS budget I was responsible for
  • 3 of 3National targets met by the end of my tenure, from three of four missed at the start
  • 6th of 135Commissioning organisations in England in the Children’s Commissioner’s detailed national assessment
  • OutstandingThe CCG’s overall NHS England rating, up from “Good”

Qualifications

  • Master’s degree in Information Technology (MSc)Newcastle University
  • Postgraduate Diploma in Management (PGDip)Birkbeck, University of London