AI Engineer

Trytuza — United Kingdom · Posted ~2 hours ago

Hybrid Visa Sponsored

Skills

Multi-agent systems AI agent orchestration AI evaluation AI auditability Trustworthy AI systems AI agents Evaluation frameworks

🔓 Log in to save this job, tailor your resume & track your apply process — 7 days free, no card needed.

Log in to add to target list

Summary ✨ AI‑Generated

Join an early-stage AI team building multi-agent systems that automate complex knowledge-work processes. You will own agent orchestration, system auditability, and evaluation frameworks that make autonomous workflows reliable and trustworthy. The role offers meaningful technical ownership, close collaboration in a growing team, and visa sponsorship for suitable candidates.

Highlights

Early-stage AI engineering opportunity focused on building multi-agent systems, with ownership of orchestration, auditability, and evaluation. Visa sponsorship is available for suitable candidates.

Description

Build the multi-agent systems that automate work banks currently do by hand. You'll own the orchestration, the auditability, and the evals that make our agents trustworthy. Welcome Thanks for taking the time to consider Tuza. When joining a company at the early stages of its growth story, there's often more than meets the eye. This is our opportunity to help you understand why your next role should be with us. We're based in Shoreditch, spend three days a week in the office, and can sponsor visas for the right candidate. Diversity is a priority for us. We build a better product when the people building it don't all think the same way. If you're still unsure at the end, drop into our offices in Shoreditch and see it for yourself. What we do We build AI agents for banks, taking on the manual work that stops them from growing. Every business that comes to a bank for a financial product gets passed from team to team, sales to onboarding to activation, and so on. Right now, every one of those teams is only ever as fast as the number of people in it. Our agents do that work without the ceiling. To be trusted with it in a bank, they have to be accurate, quick, and accountable, with auditability, explainability and security built in rather than an afterthought. That's what makes banks the hardest place to attempt this, and the place where getting it right is worth the most. Where we're at We've raised $7 million from Northzone, Connect Ventures and TriplePoint, and work with acquiring banks including Worldpay, Barclaycard and NatWest. We're earlier than that list makes us sound. The problem is real, and banks are already paying us to solve it. We're still experimenting with the product that best surfaces our value, and that's a large part of this job, rather than something that will be handed to you finished. Doing that work is a small but driven team of 12 based in our Shoreditch office who care about their work and its impact. There's also something you might already know us for. We built the UK's first payments price comparison platform, helping small businesses compare providers. It still runs today on MoneySuperMarket, Uswitch and money.co.uk. You won't be working on it, but it's why we know this market so well. How we work Extreme ownership: you own the outcome of your work, not just the output. If you see something broken or missing, you have the freedom and the responsibility to fix it.Learning velocity: curious, adaptable, and quick to pick things up. Hard problems require grit, and we embrace them.Customer-centric: we spend a day a week inside our partner banks, watching problems happen before they're even reported. Product decisions get made in that room, not in a roadmap meeting.Open and direct: we communicate openly, directly, and with intent. What you'll do Develop and optimise AI agents for live banking workflowsDesign review gates, fail-safes, and fallback behaviourBuild the observability and feedback loops behind agent performanceSpend time with our partner banks in their offices, and turn what you see there into product changes What we look for We're hiring mid- to senior-level individual contributors with at least 4 years' experience. Exceptional Python and experience orchestrating LLMs with tools like Temporal or LangChainAI automation shipped and running in productionA record of making agents reliable and efficientFluency with data pipelines and integrations We're an early-stage and intentionally lean. That means: We judge impact by the outcome of work, not the outputPriorities shift when we learn something new, and we try to communicate why, but it happensYou'll be given a direction, and the plan is yours to work out This suits people who are comfortable with ambiguity, ask early when context is missing, and would rather ship something good enough fast than perfect something late. It's less suited for those who prefer clear hierarchies, detailed briefs, or predictable roadmaps. That's not a character flaw; we just don't offer that yet. Probation is a mutual check-in. We give feedback as we go, and we expect the same in return. If the fit isn't there, we'd rather know at month two than month six. Company Benefits Generous equity.A hybrid working approach. We're based out of our Shoreditch office, with Tuesdays, Wednesdays and Thursdays as core collaboration days. The remaining days are focus days - work from where you work best.Flexible working hours. Most people arrive between 8:00 and 10:00, and leave between 6:00 and 7:00.Team trips around the world (we've been to Morocco, Istanbul and Switzerland)Teammates who really care about doing good work.