SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Zopa is building AI Banking—an intelligent, conversational experience that lets customers manage their finances through natural language. The company has developed an internal agentic platform (MCP-based tools, memory, generic frontend, evals framework) that enables rapid iteration: new customer experiences often require only prompt changes, tool definitions, and evaluation sets rather than months of engineering work.
As an AI Product Builder, you will own customer problems end-to-end: identify opportunities, decide what should exist, build it, ship it, and learn from production outcomes. This is a hybrid role between product and engineering, requiring hands-on execution rather than requirements documentation.
Key responsibilities:
- Own complete customer experiences on the agentic platform, building through prompts, tools, and configuration
- Prototype ideas yourself rather than writing specs for others to build
- Design evals that define success and inform iteration and release decisions
- Work directly in the codebase, raise PRs for engineering review, and partner with engineers on deeper technical work
- Shape agent performance in production, balancing context, cost, latency, and model choice
- Review real customer conversations and iterate improvements within days
- Lead analysis and research: interrogate data, run numbers, review transcripts to inform decisions
- Make sound judgements on risk, compliance, and customer harm, working closely with Risk and Compliance partners
You will work in a small team with high autonomy, shipping fast in a heavily regulated industry. Progression is based on the quality and impact of what you create, not team size or reporting lines.
Requirements:
- Demonstrated ability to personally build and ship customer experiences using AI; can clearly explain your decisions
- Strong product taste: know what good looks like and what is not ready to ship
- Ability to uncover meaningful customer insights and turn them into intuitive solutions
- Regular use of coding agents with thoughtful understanding of where they add value and fall short
- Experience designing evals and explaining how they identified issues traditional testing would miss
- Understanding of how LLM systems work in practice: context management, tool design, prompting, non-determinism, cost/latency trade-offs
- Systems thinking: when encountering recurring problems, solve the underlying cause
- Deep curiosity, open-mindedness, and comfort adapting as technology evolves
Bonus experience:
- Banking, fintech, payments, or other regulated environment background
- Building and operating agentic systems in production with real users and meaningful consequences
- Model routing, context engineering, or cost optimization at scale
- Voice, image-based, or multimodal customer experiences
- Side projects, open-source contributions, or personal builds
