Technical due diligence & hiring
London, UK
Taking enquiries for early 2027
AlexWilliams — Technical due diligence & engineering hiring for investors and the startups they back

I help investors find out what a startup has actually built, and whether its AI does what the pitch says. And I help founders hire engineers before they have a senior one of their own.
Plain-English reports from a lawyer turned CTO. One engagement at a time, alongside my job as CTO at Propelr.
hello@alexwllms.com- Law degree at Kent, top of the year
- First
- Insurtech valuation, employee 20
- $1.25bn
- Engineers interviewed in six months
- 50+
Services
Two ways to hire me
One is for investors, the other for founders. Scope and price are in each.
Technical Due Diligence
For investors. I look at what a startup has actually built, and whether its AI does what the pitch says.
Technical Due Diligence
For investors. I look at what a startup has actually built, and whether its AI does what the pitch says.
AI has made it cheap to build something that demos well. It hasn’t made it any cheaper to build something that lasts. A codebase written mostly with AI tools can look tidy and still be hard to change, thinly tested, or understood by one person. None of that shows up in a demo or a data room.
I read the code, talk to the engineers, and check the AI claims against what is actually running. Is it their own model, or a call to someone else’s? Do they know when it gets things wrong? What does each use cost them? Then I write it up.
I was a lawyer before I was an engineer. The report says what I found, how much it matters, and what I’d want fixed before or after you invest. I’ve sat on both sides: three and a half years inside a VC fund’s product team working with its portfolio founders, and as a CTO taken through investor technical diligence.
Sometimes I’ll stay on to help the team fix what I found.
What you get
- A written report in plain English, findings ranked by how much they matter
- A call to go through it, with the founders too if you want
- What I’d want fixed before you invest, and what can wait
Seed · from £5,000
Series A · typically £8,000–£12,000, quoted per engagement
Usually two to three weeks
Hiring When AI Writes the Code
For founders without a senior engineer yet. I build your interview process, run the final interviews, then hand it over.
Hiring When AI Writes the Code
For founders without a senior engineer yet. I build your interview process, run the final interviews, then hand it over.
Ninety per cent of developers now use AI tools. A take-home or a coding test mostly tells you how well someone uses them. Most interview processes were built before that, and a founder without a senior engineer has nobody to notice what they’ve stopped measuring.
At Marshmallow I was the Bar Raiser, the person responsible for holding the hiring standard. I ran over fifty interviews in six months while engineering grew more than fivefold. For you, I design the process, write down what good looks like at your company, and run the final-stage interviews myself. Once you have a senior engineer who can hold the bar, I hand it all over.
I don’t find candidates. You bring them, and I make sure you can tell which ones to hire. If I’ve already done diligence on your company, I know the codebase, which saves us both time.
What you get
- An interview process that still works when candidates use AI
- A written bar: what you expect at each level, and how you’ll tell
- Final-stage interviews run by me until someone on your team can take over
£6,000 · process design and up to eight final interviews
£400 per final interview after that
Selected work
What actually changed
These are jobs I’ve held, not client engagements. Every number here is one a reference call would confirm.
3 yrs
On the original architecture, no rebuild
Propelr
CTO & technical co-founder
Proptech · pre-seed to seed+ · 2023–now
Situation
A pre-seed proptech with a product thesis, no engineers, no platform, and enough runway to build one version of it.
What I did
- Built the entire technical platform from scratch and took it through to a seed-plus stage business, making the architecture and vendor decisions that a company at that size cannot afford to get wrong twice.
- Rolled AI tooling out across the engineering team — choosing what to adopt, where it genuinely helped, and the review practice around code nobody typed by hand.
Result
The platform carried the company from pre-seed to seed-plus without a rebuild. I am still the CTO, which is the strongest thing I can say about the decisions.
10→50+
Engineers, in under two years
Marshmallow
Engineer → Senior Engineering Manager
Insurtech · employee 20 · 3 years
Situation
A car insurance startup heading for the round that would value it at $1.25bn, with under ten engineers and no engineering management function at all.
What I did
- I became the company's first engineering manager, then ran two teams and line-managed fifteen engineers plus a manager responsible for ten more.
- The founders chose me as the company's Bar Raiser: over fifty interviews in six months, final-stage culture interviews on their behalf, and the first backend engineer, frontend engineer and technical lead hired for a new Budapest office.
Result
Engineering went from under ten people to over fifty without the bar dropping, and I hired and coached the third and fourth engineering managers who took it from there.
12 moTo build the function and hand it overVita Mojo
12 mo
To build the function and hand it over
Vita Mojo
Senior engineering manager
Hospitality tech · Series B · 1 year
Situation
A Series B business with engineers in London and Moldova, no engineering management function at all, and a raise to deploy on growing the team quickly.
What I did
- I built the management function outright — 1:1s, objectives, review and a levelling framework across seven cross-functional teams.
- I rebuilt their hiring loop on the Bar Raiser model, and coached a junior engineering manager running a team of sixteen in the Moldova office.
Result
They finished with a working management layer and a hiring process their own people could run without me. That was the brief: build something that does not need the person who built it.
£500kA year in third-party feesMarshmallow
£500k
A year in third-party fees
Marshmallow
Senior engineering manager
Insurtech · Efficiency team · 2021
Situation
The business was paying a third party to run its monthly instalments, on a contract with a hard exit in November 2021 and no replacement built.
What I did
- I led the Efficiency team to remove Premfina and launch our own monthly instalments product against the deadline.
Result
Shipped on the deadline with no issues in the first thirty days, and took about £500k a year off the fee line.
About
Who you would be hiring
I came to engineering from law. I graduated top of my year at Kent with a first, practised in Guangzhou for eighteen months, then taught myself to code.
I was the Money Advice Service’s first junior engineer on a team that grew past thirty, then spent three and a half years as the in-house product team for Forward Partners, an £81m venture fund, building and shipping for its early-stage portfolio. I joined Marshmallow as employee 20 and stayed through the round that valued it at $1.25bn — first as an engineer, then as its first engineering manager, and finally running two teams and acting as the company’s Bar Raiser while engineering went from under ten people to over fifty. Vita Mojo hired me after that to build their engineering management function outright.
Since 2023 I’ve been CTO and co-founder at Propelr, a UK property transactions platform. I built the platform from pre-seed to seed-plus and brought AI tools into how the team works, so I see what AI-written code does to a real codebase every week. The law is also why diligence suits me. It’s the same job as legal advice, gathering evidence and writing something a person can act on, in a different field.
Where that happened
Propelr
CTO & technical co-founder
Built the platform from pre-seed to seed-plus, and rolled AI tooling out across the engineering team.
2023–now
Vita Mojo
Senior engineering manager
Hired to build the engineering management function outright, and rebuilt hiring on the Bar Raiser model.
2022–23
Marshmallow
Engineer → senior engineering manager & Bar Raiser
Employee 20. Engineering went from under ten to over fifty, through the round that valued the company at $1.25bn.
2019–22
Paws.com
Frontend engineer, then head of product delivery
Shipped the build-a-box proposition against an eight-week deadline.
2018–19
Forward Partners
Frontend engineer
The in-house product team for an £81m fund, building across its early-stage portfolio.
2015–18
Before that: Money Advice Service, 2013–15. Before that, a first-class law degree at the University of Kent and eighteen months practising foreign direct investment law in Guangzhou.
Writing
How I think about this
What I look for before an investor backs a company, and how to hire engineers when the tools write the code.
Is the AI real? What I ask before an investor backs it
Nearly every pitch now mentions AI, and the word covers very different businesses. These are the questions I ask to find out which one you’re looking at.
6 October 2026
The four technical decisions I’d check before investing
Most of what a startup’s engineers argue about can be undone without much pain. Four decisions can’t, and they rarely come up in a pitch. They’re where I look first.
6 October 2026
Your interview loop is testing a skill that no longer exists
Take-homes are dead and live coding is contaminated. Most teams have quietly stopped trusting their own process. This is how I’d rebuild it.
4 August 2026
Reference
What the people who hired me say
Alex was our first engineering manager and helped us grow the team to six engineering managers and 60 engineers. He acted as a Bar Raiser for hires across the entire company as we scaled to our billion dollar valuation. He led the delivery of our first non-motor insurance product, home insurance, assembling the team from scratch to deliver a finished product within six months.
Co-founder & Chief Architect · Marshmallow
Questions
Questions I get asked on the call
You’re already a CTO. How does this work?
You’re already a CTO. How does this work?
I’m CTO at Propelr and I don’t hide it. It’s why I take one engagement at a time. It’s also why my view of AI-written code is current: I’m dealing with it in my own team this month. If you need someone in your office every day, I’m the wrong person, and I’ll say so on the call.
Do I need an AI specialist instead?
Do I need an AI specialist instead?
If the startup trains its own models, part of the question needs an ML specialist, and I’ll tell you that at the start. Most don’t. Most build on someone else’s model, and the questions that matter are engineering ones: does it work on real data, do they know when it’s wrong, what does it cost to run, and could they switch provider. Those I can answer.
What does the report look like?
What does the report look like?
It’s written for someone who doesn’t read code. A short summary of what I found and how much it matters comes first, then the detail behind it. It’s a technical report, not a legal opinion. I have a law degree, but I’m not a practising lawyer.
How do you handle conflicts?
How do you handle conflicts?
I won’t take paid work from a company while I’m assessing it for an investor. I won’t do diligence on anything that competes with Propelr. And I won’t do diligence on a company I hold equity in. If something else might look like a conflict, I’ll tell you before we start.
Do you stay on afterwards?
Do you stay on afterwards?
Sometimes. Once the report is with you, if the founders want help fixing what I found, I’ll stay on for a while. If the team needs to grow, that’s the hiring work.
What don’t you do?
What don’t you do?
Financial diligence, legal diligence or security penetration testing. Those need specialists, and I’ll point you to one. I don’t find candidates for hiring work either; you bring them. And I won’t soften a finding because the round is busy or the founders are likeable.