Software Engineer (Mid / Senior / Staff / Principal)

Most organisations make their biggest decisions — what to build, how to price it, who it's for — on a thin slice of evidence. A handful of interviews, a survey that took six weeks to field, the loudest voice in the room. The teams that can afford to do better spend hundreds of thousands on research that's often stale by the time it lands.

Electric Twin is a behavioural simulation platform that changes the economics of understanding people. We build synthetic populations from real customer data, grounded in behavioural science, so teams can stress-test pricing, product, and strategy decisions in hours rather than quarters.

The challenge is making it trustworthy — that's the bar we've set. We combine foundation models with rigorous data science, validated across real populations, new questions, and high-stakes decisions. It also needs to be explainable, so customers understand the outputs. That validation is what makes it credible — without it, it's just faster guesswork. We've already run over 40,000 evaluations of populations covering 155 countries to train our prediction engine. Academic research conducted with Professor Michael Muthukrishna at the London School of Economics found our technology delivers insights 10,000 times faster than traditional research methods, with 95% accuracy. The world's leading companies and governments already use Electric Twin to make better decisions.

AI is part of what we build, but it isn't the product. The hard work is system design, orchestration, and making outputs trustworthy enough that people act on them.

Done well, this democratises something that's been gated by budget for too long. The depth of customer understanding that used to take a quarter and six figures becomes available in an afternoon — usable on every decision, not just the ones that justify the cost.

Electric Twin has raised $14m in funding, including a $10m round led by Atomico, backed by LocalGlobe, Mercuri, and Samos Investments. Our angels include Marc Andreessen; Cal Henderson (co-founder and CTO of Slack); Eric Salama (former Kantar CEO); and Tom Shinner (COO of Entrepreneur First). This follows a previously undisclosed $4m pre-seed.

Electric Twin was founded by Dr Ben Warner, a physicist and former Chief Adviser on Digital and Data to the Prime Minister, and Alex Cooper, a former military commander who established the UK's mass testing response during the pandemic. Having worked together in Number 10 at the heart of the COVID response, they founded Electric Twin so decision-makers had the tools they needed to make better decisions.

The work at Electric Twin is about turning complex decision-making into usable software. Engineers build the systems that let customers move from a broad question to a structured, trustworthy outcome — combining product experience, workflow logic, backend services, and platform capability into something coherent.

Some of that work is highly customer-facing: shaping interfaces that make sophisticated workflows feel clear and accessible. Some is deeply systemic: designing the services, data models, APIs, and execution layers that make those workflows reliable and reusable. And some sits at the boundary between science and product, taking capabilities validated in research and turning them into robust product behaviour customers can depend on.

The common thread is ownership of the whole problem, not just the implementation. Engineers here help shape the capability, decide what trade-offs are worth making between speed and rigour, judge when something is ready for customers and when it needs another pass, and stay with it as it evolves. Specs aren't handed down — they're worked out by the people building the thing.

You'd be joining a delivery team of twelve: five engineers, five data scientists, and two designers, reporting to Rob Cherry, VP of Engineering. The shape of the team is deliberate — the science isn't a separate function that hands things over; it sits inside the team building the product.

Our stack is Python, FastAPI, TypeScript, React/Next.js, PostgreSQL, GitHub, and Google Cloud Platform. Prior experience with these specific tools isn't required — we're happy to support your learning where you have less experience. Our most recent Principal Engineer hire came predominantly from a Java background, even though there's no scope to use Java in the role.

We don't have a fixed checklist of years, frameworks, or job titles. The work is what matters, and the people who do it well tend to share three things.

They take ownership of problems, not just tasks. The work here often starts ambiguous — a customer need that isn't fully formed, a capability that needs shaping into a product, a trade-off no one has made yet. We're looking for engineers who see that as the interesting part, rather than something to be resolved by someone else before they begin.

They have good judgement under uncertainty. Most of the decisions that matter are trade-offs: speed vs rigour, scope vs polish, building now vs validating further. We're less interested in whether you reach exactly the same answer we would, and more in whether you can reason about the trade-off clearly and explain why you landed where you did.

They ship and learn. Thinking carefully about a problem is necessary but not sufficient. The engineers who thrive here are the ones who get things in front of customers, learn from what happens, and iterate — rather than waiting for the design to be perfect before anything moves.

A note on experience. We've hired engineers across a wide range of experience levels and stacks, and the pattern is clearer than we expected: years matter less than what you've actually owned. People thrive here at every stage — what they share is a track record of having taken real responsibility for real problems, whether that came after two years or twenty. The ones who tend to struggle are those who've mostly worked inside tightly defined roles where the hard decisions were made elsewhere — not because they aren't capable, but because the work demands a kind of ownership that usually has to be learned by doing it.

If you've owned something meaningful end to end — shaped it, made the trade-offs, shipped it, and lived with what happened next — we'd like to hear from you, regardless of how long that took.

Small team, experienced and low-ego, with fast feedback and few layers. People care about getting to the right answer rather than being right — founders included. Disagreement is welcome and decisions are explained.

We're based in Soho, central London, and spend most of our week together in the office. Collaboration here is real, not ceremonial — the work genuinely improves when people are in the room.

Plenty of companies will tell you they value work-life balance while filling your calendar with 7pm "optional" meetings and stocking the kitchen so you never leave. We're not that. People here have kids, partners, dentist appointments, school plays, and the occasional delivery that somehow needs a signature in the middle of the day. The expectation is simple: do excellent work, and have an actual life alongside it.

• Equity. All full-time employees are enrolled in our share options scheme at the end of probation — explained in detail when you join.

• Pension. We match your individual contributions up to 8% (16% total) under a net pay structure.

• Private healthcare with BUPA.

• Unlimited leave — and we genuinely expect you to take it.

• £1,000 annual L&D budget to spend as you see fit.

• Cycle to Work Scheme.

• Kit. Laptop plus any accessories you need (monitor, keyboard, etc.) to do your best work.