Job Description

AI Product Manager, Open Agentic

Open Agentic is where we're building the next generation of AI products for insurance — products that can reshape how insurance is distributed, understood and managed. Through partnerships with brands including Avios and MoneySuperMarket, we already have AI products live in the market with household names.

We're hiring two AI Product Managers to help us 10x the pace. You'll work directly with the Director of Product, Open Agentic, on some of the highest-profile products in the company, with significant autonomy and end-to-end ownership. You'll build to prove an idea with working prototypes, then work with our engineering team to take it into production.

Think startup within a startup: high agency, high visibility, very little bureaucracy and lots of room to grow.

What you'll do

This is a hands-on product role — as much about deciding what to build as building it. You'll work across product thinking, prototyping, engineering, AI, UX and deployment rather than owning one narrow part of the stack.

Current Open Agentic product areas include:

• Global Product Graph, our structured intelligence layer for insurance products.
• Companion, our consumer-facing AI agent. Today it guides customers through buying insurance. Before long, it should let them put their insurance on autopilot.
• AI distribution, including ChatGPT Apps, MCP quote integrations and partner deployments.
• Agentic Safety Platform, which helps companies evaluate, monitor and manage customer-facing AI deployments.

We don't expect these to be the only things you'll work on. Part of the role is helping us discover and build what comes next.

You'll:

• Take ambiguous product ideas and turn them into working software.
• Decide what to build now, what to defer and what not to build at all.
• Ship quickly, get things in front of real users and iterate on what you learn.
• Own products beyond the prototype — from zero to one, then from one to 100.
• Talk to customers and partners to find the problem worth solving, not just to validate a solution.
• Own the Global Product Graph — the ontology, the data, and the workflows that keep it accurate.
• Develop the insurance intelligence on top of it: product scoring, compliance evaluations, and what "good" means for a regulated AI deployment.
• Help shape both what Open Agentic builds and how we build it.

Within your first three months, we'd like you to have taken meaningful ownership of an AI-native product and shipped it. Within six months, to be owning a major product or workstream end-to-end and taking it towards real customer adoption and revenue.

Who we're looking for

We're looking for builders — people who get excited by a blank page, move quickly and naturally take ownership when something needs doing.

You'll probably have an engineering or computer science background and be able to build things yourself, but your ambitions sit in product: more interested in whether we're building the right thing than in owning the system that runs it.

You're probably more excited by the product and commercial applications than by perfecting the engineering, and you want the next few years to grow your product and commercial mind rather than your engineering craft.

You can build. You should be able to show us something you've personally built and deployed — we don't care whether it has 1 user (you) or 10 million — and explain how it works and the trade-offs you accepted. We're not asking you to own production systems. We are asking that you can prove an idea without needing a team.

You have product judgment. You start from the problem and who has it, not the feature. You're opinionated about what to build, comfortable arguing for what to leave out, and you can say how you'd know whether it worked. You notice when something feels bad — good UX, design and clear communication matter to you alongside good engineering — and you care whether people actually use the thing, not only whether the

You're AI-native. AI isn't something you recently added to your workflow because everyone else is talking about it — it's already fundamentally changed the way you build. You're deeply comfortable with tools such as Claude Code, the latest coding models, agentic development workflows and platforms such as Vercel, and you have opinions about which are good at what. You've also built with models rather than just alongside them, and know where they're reliable and where they aren't. You keep up with new models and research because you find the space genuinely exciting, and you have a view on where it's going.

You have high agency. You don't need a perfectly written PRD, detailed Jira tickets and three meetings before you can start. You're comfortable with: "Here's the problem. Go figure out what we should build."

You're hungry to grow. You'll probably have 2+ years building products, but we're far more interested in what you've done than how long you've worked or what credentials you have.

Nice to have

None of this is required, but any of it is useful: insurance, financial services or another regulated industry; shipping AI or ML products in production; evaluations, model judging, red-teaming or guardrails; knowledge graphs, structured extraction or retrieval; Python, TypeScript and modern cloud infrastructure; or fast-moving startup experience.

Why join?

We have AI products live with household-name brands, real customers using them and the regulatory groundwork in place to deploy AI safely in insurance.

What we want next is pace — more products, shipped faster, on foundations that already exist.

You'll get:

• Real ownership: take products that matter to Open's future from an idea to something customers pay for.
• Direct access to leadership: work with the Director of Product and closely with senior leadership, on work seen across the company and by major external partners.
• Speed: a small team where good ideas become live products quickly, using the best available AI tools rather than legacy ways of working.
• Room to grow: as Open Agentic scales, your scope should scale with it.
• Competitive compensation, including share options, so everyone shares in our success.
• Flexible working, paid company parental leave and a personal development allowance.

If you want a narrowly defined role where somebody else decides what gets built, this probably isn't the right job.

If you want to build things from scratch, ship constantly and have an outsized impact on an AI business that's already live and about to get a lot bigger, we'd like to meet you.

Role location

Based in our Marylebone office in London, hybrid, with at least three days a week in the office, a rhythm that supports collaboration and the fast, high-trust culture we're building.

You'll work with colleagues across the UK and Australia, including our broader Product and Engineering teams.

You'll need the right to work in the United Kingdom.

How to apply

Send us your CV, but don't write a cover letter. Instead, answer three short questions:

1. Send us a link to something you've built and deployed recently — a side project, website, tool or something professional. In 20 words or fewer, what is it and what did you personally do?
2. What's your current preferred coding model / AI development setup? Tell us why in 20 words or fewer.
3. One AI product or tool that you love — not ChatGPT or Claude. Why, in 20 words or fewer.

Don't overthink the answers. We're just looking for genuine builders with curiosity and opinions.

Interview process

We want the process to be fast, practical and representative of the actual job. There's no algorithm trivia and no whiteboard coding test.

1. Initial conversation with the Director of Product, Open Agentic — what you've built, how you work and what you're looking for next.
2. A take-home build. We'll give you a real Open Agentic product problem and ask you to build and deploy something, using Claude Code or whatever tools you like. Then you'll walk us through it with a small group from product and engineering.
3. Final conversation with Open's co-CEOs, focused on ambition, culture and where you want to grow.

We want to move quickly when we find exceptional people.

More about us

Insurance is one of the world's least digitally mature industries: confusing, paper-based and heavily intermediated. We launched in Australia in 2016 to digitise the entire process, opened our London office in 2022 and our Manchester office this year. We're a Certified B Corporation with a team across the UK, Australia and New Zealand.

We do this through three products. Open Embedded lets businesses offer insurance directly inside their own apps and websites. Open Agentic uses AI safely and responsibly to guide customers through buying a policy, answering questions in plain English. Open Intel spots the moments when someone is most likely to need cover — like buying a car or moving house — and helps us reach them at the right time.

Diverse perspectives and lived experiences make us better at what we do. If your CV doesn't tick every box, apply anyway. Let us know if you need any adjustments through the process and we'll make it work.

Open is for builders — people who are intellectually curious, courageous and reject the idea that "it can't be done". If that sounds like you, apply and let's build the future of insurance together.