AI Product Manager

The Citation Group; Wilmslow; Hybrid; GBP 90,000 - 150,000; Full time

Job summary

The Citation Group is seeking an AI Product Manager to lead one of our Compass or Tiger teams. You will work hands-on with engineers, using agentic coding tools to prototype and ship AI-native products that deliver real client impact.

You will define and execute the product strategy, measure eval results, and track adoption, revenue impact and cost to serve, ensuring the work moves from concept to production with rigor and speed.

Qualifications

• Proven ability to ship AI products to production.
• Experience with evals to measure AI feature success.
• Fluency in product analytics and business impact.
• Hands-on with AI tooling and prototype development.

Responsibilities

• Lead a Compass or Tiger team from problem definition to live product.
• Set day-to-day direction based on client outcomes and impact.
• Hands-on coding with engineers to prototype and move fast.
• Measure commercial value and track adoption, revenue impact, and cost to serve.

Skills

Product builder

AI product shipping

Agentic tooling

Product analytics

Commercial mindset

Evals expertise

Code collaboration

Claude Code

Tools

Claude Code

Job description

Location: Hybrid- typically 3 days office

Reports to: Product Director

Overview

Citation Group helps over 120,000 SMEs across the UK, Australia and Canada take care of the non-negotiables of running a business — HR and employment law, health & safety, ISO certification, background screening, cybersecurity, e-learning and more. Our mission is simple: build better businesses to create a better world.

We're changing how we build products, not just what we build. Small, high-leverage teams — we call them Compass or Tiger teams — pair a product manager with engineers using agentic coding tools and a quality/evals specialist, and ship AI-native products and experiences faster and to a higher bar than the traditional model allows. This is the frontier of Citation's product organisation: fewer layers, more building, and AI woven into both the products we ship and how we ship them.

As AI Product Manager, you'll lead one of these teams. This isn't a roadmap-and-backlog role — you're a builder. You'll have used agentic coding tools yourself, built evals to know whether an AI feature is actually good, and shipped AI products into production, not just prototyped them. You'll take a Compass or Tiger team from problem to live product, working hands-on alongside your engineers rather than translating requirements at a distance.

Why this role? Why now?

• Build genuinely AI-native products. Not bolt-on AI features — agents, intelligent workflows and experiences that anticipate and act on a client's behalf. You'll define what good looks like as much as build it, because much of this is new ground for the industry as well as for us.
• Lead from the front, not the roadmap. You'll run a small, high-leverage team and be closer to the code and the evals than a traditional PM — using agentic tooling daily, shaping what gets built and how, and proving the pod model works (1 product, 2 AI Engineers).
• A defined mandate on a live initiative. You'll be shipping into one of Citation's highest-priority AI programmes, with real client impact and real visibility across Product and Engineering leadership.
• Be part of redefining how Product works. Citation is actively reshaping its product operating model around agentic ways of working. You'll be one of the practitioners proving it out, not a passenger to a change someone else designed.

What you'll be doing

Own the team and the outcome

• Lead a Compass or Tiger team — a small pod of engineers using agentic coding tools and a quality/evals specialist — from problem definition through to a live, working product.
• Set direction for the team day to day: what gets built, in what order, and why, grounded in client outcomes rather than output volume.
• Work hands-on alongside your engineers, using agentic coding tools yourself to prototype, unblock and move fast.
• Assess the commercial value of what you're building before and after you build it — sizing the opportunity up front, then tracking adoption, revenue impact and cost to serve once live — so the team's time goes on what actually moves the needle commercially, not just what's technically interesting.

Build AI-native products, not SaaS features with AI bolted on

• Design product experiences that anticipate and act on a client's behalf, rather than simply presenting information for a human to act on.
• Define and build evals so the team knows whether an AI feature is actually working — not just whether it shipped.
• Take AI products through to production, including the unglamorous parts: reliability, cost, latency, guardrails and what happens when the model gets it wrong.

Discover and validate with real clients

• Run discovery directly with clients to understand where an intelligent, proactive product genuinely earns its place versus where it's AI for its own sake.
• Own the product analytics for your team's product — usage, adoption, retention, funnel and cost/performance metrics — and use it alongside client feedback and evals results to decide what to build next.

Work across Product and Engineering

• Partner closely with the wider engineering teams on tooling, models and platform capability.
• Bring back what you learn — on agentic ways of working, on evals, on what AI-native product development actually takes — to the wider Product team.

What we're looking for

• A product builder first. You can write a clear requirements doc when the situation calls for it, but you're just as comfortable in the codebase, prototyping alongside engineers using agentic coding tools (Claude Code or equivalent) as a regular part of how you work, not as an experiment.
• Real experience shipping AI products to production. Not a pilot that stayed a pilot — a live product, with real clients, that you built and are accountable for.
• Hands-on with evals. You know how to define what “good” means for an AI feature and how to measure it, and you've built or used evals frameworks to do so.
• Strong on product analytics. You're fluent in the numbers — usage, adoption, retention, funnel drop-off, cost and performance metrics for AI features — and you use them, alongside evals results, to decide what the team builds next rather than relying on instinct or anecdote.
• Commercially switched on. You can build a business case, size an opportunity and track it through to delivery — you treat commercial return as your responsibility, not something Finance checks after the fact.
• Comfortable with ambiguity. AI-native product development doesn't have a settled playbook yet. You're happy defining the approach as you go rather than waiting for one to be handed to you.
• Strong product fundamentals. Discovery, prioritisation, working with data and clients to decide what matters — the core PM craft still applies, it's just applied faster and closer to the build.
• B2B SaaS experience preferred, ideally in a multi-brand or multi-business-unit environment.

What's in it for you

• The chance to build genuinely AI-native products from the ground up, inside one of Citation's most consequential initiatives.
• A hands-on, builder-focused role at the leading edge of how AI is reshaping product development.
• Visibility across Product and Engineering leadership, with real ownership of what you ship.
• A supportive culture that prizes action, innovation and getting great things done — for our clients and each other.