Technical Product Manager

Prism Digital; Greater London; Hybrid; GBP 75,000 - 90,000; Full time

Benefits offered by this job

Equity component

20 days annual leave

Wednesdays from home

Direct access to founders

Job summary

Prism Digital is seeking a Technical Product Manager to drive AI evaluation and product quality for an early-stage platform in central London. You will own both product quality—evaluation suites, graders, pass rates—and product ownership—PRDs, prioritisation, delivery—from idea to shipped features.

The role requires hands-on AI work, collaboration with engineers, and close interaction with customers, with four days in Victoria and Wednesdays from home.

Qualifications

• 3–5 years in Product or a closely related role owning outcomes and working with engineers.
• Hands-on technical grounding: roughly a year building software or meaningful AI workflow experience.
• Day-to-day use of AI tools to solve real problems, not just experimentation.

Responsibilities

• Define what good performance looks like across the product.
• Build and tune AI evaluation suites, test scenarios and graders.
• Track pass rates and regressions, close gaps.
• Turn customer feedback into prioritised product problems.
• Own features from problem definition through delivery and validation.

Skills

Product ownership

AI experience

Data-driven

Engineering collaboration

Customer-facing

Education

Strong academic background

Tools

Linear

Notion

Slack

Job description

Technical Product Manager | AI Evaluations, Product Quality, Agentic Workflows | Early-Stage AI Platform

• Salary: c. £80k base, plus equity. Some flexibility around the overall package
• Location: Victoria, central London. 4 days a week in the office, Wednesdays from home
• Right to work: existing UK right to work, no sponsorship
• Experience: typically 3 to 5 years in Product or a closely related role

I’m working with an early-stage AI startup building an assistant for property managers, letting agents and Build-to-Rent teams. It handles communications, maintenance, compliance and scheduling, and it’s already live with pilot customers managing thousands of properties.

They’re around 25 people, including roughly 13 engineers, and they’re now hiring a Technical Product Manager to sit between Product, Engineering, AI and customers.

About half the job is owning product quality: evaluation suites, test scenarios, graders, pass rates and the daily investigation of things the AI has got wrong. The other half is Product ownership: customer problems, PRDs, prioritisation and taking features through delivery.

Fair warning, this isn’t a purely strategic roadmap role. There’s plenty of influence, but you’ll also be in the detail working out whether an issue is the product, the data, the configuration or the user.

What They’re Looking For (Non-Negotiables)

• 3 to 5 years in Product, or a closely related role where you owned outcomes and worked directly with engineers
• Hands-on technical grounding: roughly a year building software, or meaningful experience building agentic AI workflows yourself
• Genuine daily use of AI tools to solve real problems, not just occasional experimentation
• Able to reason credibly with engineers about systems, feasibility and technical trade-offs
• Strong instincts for product quality, with the rigour to measure it rather than rely on gut feel
• A methodical, root-cause approach to investigating problems
• Comfortable with data, dashboards and operational tooling
• Strong with customers as well as technical teams
• High ownership and comfortable in an early-stage company
• Strong academic background
• Existing UK right to work
• Comfortable being in Victoria four days a week

They may consider someone closer to two years if the technical and AI experience is unusually strong.

What You’ll Own

• Define what good performance looks like across the product
• Build and tune AI evaluation suites, test scenarios and graders
• Track pass rates, spot regressions and drive the work to close gaps
• Use real customer conversations and calls to create better test scenarios
• Run the daily critical-error review and investigate flagged conversations
• Separate configuration, data and user issues from genuine product bugs
• Turn customer feedback into clearly prioritised Product problems
• Own features from problem definition through delivery and validation
• Write PRDs and success criteria in Notion, with delivery managed through Linear
• Help launch new modules across maintenance, prospects, compliance and arrears

Nice to Haves

• Experience writing test scenarios or automated graders
• Previous ownership of QA or a product quality bar
• Python or software engineering experience
• Basic SQL or confidence exploring data independently
• Linear, Notion and Slack
• Startup or scale-up experience
• PropTech or property experience, although they’ll teach you the domain

Why Join / Projects

Testing normal software is comparatively tidy: give it an input and check the output. Testing an AI product that can respond differently to the same input is a much more interesting Product problem.

You’ll decide how that quality is measured across voice, email, maintenance, compliance and customer communications, then use the results to influence what gets built next. Quality isn’t being handed off to a separate QA function here. It is part of the Product itself, and you’ll own it.

You’ll work directly with the Product Lead, engineers, AI team and founders. The team is small enough that your judgement will genuinely affect the roadmap, with room to grow either deeper into AI evaluation or broader into Product leadership as the company expands.

Employee Benefits

• c. £80k base, with some flexibility around the overall package
• Equity component
• 20 days annual leave
• Wednesdays from home
• Direct access to the founders and Product leadership
• Significant ownership over a live AI product