The role
We’re looking for a Senior Product Analyst to join our AI Tribe. This is a hands-on, builder-style role for someone who thinks in experiments, writes code daily, and is as comfortable building a prototype evaluation framework as they are presenting findings to a product lead.
The role is technical at its core: you’ll be writing SQL and Python to dig into enrichment performance, building evaluation pipelines to test AI-driven approaches, and prototyping experiments that push the accuracy of our transaction intelligence forward. At the same time, this isn’t a data science role — you’ll be expected to prioritise ruthlessly, move fast, and care about outcomes, not just model performance.
The AI tribe
The AI Tribe sits at the heart of how Zopa turns data into intelligent products. We build and evolve the capabilities that transform raw transaction data, Open Banking data, and other behavioural signals into reliable, structured intelligence.
A core focus of the tribe is transaction intelligence — how raw transaction data is enriched, categorised, structured, and made usable across the business. This includes merchant and category enrichment, data quality and consistency, and ensuring transaction data integrates cleanly into our broader data models and customer-facing products.
As we move toward more AI-first ways of working, the tribe is also rethinking how transaction data should be accessed and used — not just by analysts and product teams, but by AI systems and LLM-powered tools. The work is foundational and long-term in nature: improving systems, shaping strategy, and building capabilities that compound in value over time.
A day in the life...
- Write SQL and Python to analyse enrichment performance across merchant types, transaction patterns, and edge cases — quantifying gaps, not just flagging them
- Form specific hypotheses about why accuracy breaks down in certain segments and design structured tests to validate them
- Build prototypes: prompt variants, alternative classification logic, lightweight enrichment rules — tested and evaluated before any engineering resource is committed
- Design and run rigorous evaluations, including LLM-as-judge pipelines, to measure whether a change actually improves things and where it introduces new failure modes
- Work directly with LLM APIs to experiment with enrichment approaches — prompt engineering, structured outputs, retrieval-augmented approaches — with a clear eye on accuracy, cost, and latency trade-offs
- Build the feedback loops that tell the team what’s working: ground truth datasets, labelling processes, evaluation metrics
- Communicate findings and trade-offs clearly to Product and Engineering, translating technical results into actionable decisions
- Stay close to the AI landscape and proactively bring in new approaches when there’s a good reason to
About you...
- Technically strong: fluent in SQL and Python, comfortable writing code to test ideas rather than waiting for someone else to build it
- You’ve worked hands-on with LLM APIs — prompt engineering, structured outputs, evaluation — either in a professional setting or through personal projects
- You think in hypotheses: you form a view before looking at the data, design a test that could falsify it, and interpret results with appropriate scepticism
- Experience designing and running structured experiments, with a solid grasp of methodology (sample sizing, control/treatment design, what makes a result meaningful)
- You’re rigorous about evaluation: you know that “it seems better” isn’t good enough, and you’ve built or contributed to evaluation frameworks that give real signal
- Naturally curious about messy, imperfect systems — you want to understand failure modes at a granular level, not just measure aggregate accuracy
- Comfortable working in ambiguity: you can scope your own problem, prioritise what to test first, and know when good enough is good enough
- Clear communicator: you can explain what you built, what you found, and what it means — to engineers, to PMs, and to stakeholders who don’t care about the implementation details
- Experience in a fast-paced, data-led environment (fintech, banking, start-up, or consulting
At Zopa we value flexible ways of working.
We value face-to-face collaboration and a good work-life balance. This hybrid role requires you to come to our London office 2-3 days a week.
You'll also have the option of working from abroad for up to 120 days a year!* But no matter where you are, we’ll make sure you’ve got everything you need to thrive, both in your work and home life, from day one.
*Subject to having the right to work in the country of choice
Diversity Statement
Zopa is proud to offer a workplace free from discrimination. Diversity of experience, perspectives, and backgrounds leads to better products for our customers and a unique company culture for our people. We are made up of nearly 50 nationalities, have a DE&I forum made up of Zopians wanting to make a difference and we are proud of our culture where everyone can bring their full self to work. Our approach to DE&I is reflected in our hiring process so please let us know if you require any reasonable adjustments.
Our approach to AI in interviews
At Zopa, AI isn't something we're testing out — it's part of how we work every day. As a proud partner of Jobs 2030, we're committed to building AI fluency across our workforce, and we expect Zopians to use AI as part of how they do their jobs.
Because of that, we want to be transparent about how we think about AI use during our hiring process.
Behavioural and competency-based interviews: please don't use AI. These conversations are designed to understand you — your experiences, your judgment, and how you've approached real situations. An AI-generated answer can't tell us that. What it can do is get in the way of us finding out whether we're the right fit for each other.
Technical interviews: it depends on the role. Some technical stages actively welcome AI use, others don't. Your Talent Partner will let you know what's expected at each stage. Where AI is part of the assessment, we'll be interested not just in the outcome, but in how you used it – the tools you chose, your reasoning, and the decisions you made along the way.
Zopa Bank London, England Office
Tooley Street, London, United Kingdom, SE1 2QG
