Generative AI Engineer

AI Engineer, AI Lab

Location: Copenhagen or London

Are you our next AI Engineer in AudienceProject's AI Lab?

AudienceProject is looking for an experienced AI Engineer to join our internal AI Lab. This is a role for someone who has already built and shipped AI systems in production — not for someone looking for a place to start learning. We want a builder who brings real knowledge of the field, inspires the people around them, and is happy to roll up their sleeves on the systems we put in front of customers.

AudienceProject is a leading SaaS company delivering audience measurement and research solutions to major advertisers, agencies and publishers across Europe. Our platform combines large-scale exposure data, panel infrastructure, deep technical integrations with global platforms such as Google, Meta and Amazon, and synthetic population modelling to help clients understand campaign reach, frequency and audience composition across channels.

About the AI Lab

The AI Lab is AudienceProject's dedicated unit for translating modern AI — LLMs, agentic systems, retrieval, evaluation — into real leverage across the company. We exist to do two things:

• Accelerate AudienceProject's internal use of AI. We evangelize, teach, pair with teams across engineering, data science, product and commercial, and give them concrete technical support so they get real productivity gains, not surface-level adoption. This covers everything from AI-assisted coding (Claude Code, Codex, Copilot) to internal agents, skills and MCP-based tooling that take the friction out of day-to-day work.
• Build AI components that ship in our products. We design and operate production AI building blocks — tools, agents, retrieval and context layers, orchestration, guardrails, evals, observability — that our product teams plug into. The AI Lab is what makes it realistic to deliver bespoke apps, dashboards and workflows faster and leaner than a traditional setup.
• The Lab is small and senior by design — initially four people. You will be one of the founding members shaping how it operates and what it delivers.

What you will do

• Design, build and operate production AI components — agents, tools, skills, retrieval, evaluation harnesses, guardrails and observability — that ship inside AudienceProject products.
• Partner directly with our product teams to embed AI capability into customer-facing apps, dashboards and workflows. You will sit close to the engineers and data scientists actually building the product, not in an ivory tower.
• Drive internal AI adoption across the company. Run brown-bag sessions, pair with engineers and data scientists, write internal playbooks, set up shared infrastructure (MCP servers, skills, evals, prompt management, cost/usage tracking) and follow up to make sure the leverage actually lands.
• Set the bar for how we build AI at AudienceProject — patterns for agent design, context engineering, evaluation, cost control, safety and production operations. You will be expected to bring opinions backed by real-world experience.
• Stay close to the frontier. Track new models, frameworks and techniques, evaluate what is useful for us, and bring it inside with concrete recommendations and prototypes.
• Help shape the AI Lab itself — how it operates, how it partners with the rest of the organization, and how it scales as we grow.

Who you are

You are an experienced engineer who is already operating at the production end of modern AI. You can hold your own in a deep technical conversation about agents, evaluation, retrieval or context engineering, and you can ship code that other engineers can actually maintain. You are excited about this field for its own sake, not just because it is fashionable, and you bring enough depth that the people around you learn from working with you.

We imagine you bring:

• Substantial hands-on experience designing, shipping and operating production AI / LLM-based systems — not experiments and notebooks, but systems with real users, real cost budgets and real failure modes.
• Strong software engineering fundamentals: Deep understanding and experience with the finer and less obvious problems of software design and architecture, such as extensibility, security, availability and organisational fit.
• Deep working knowledge of the modern AI stack: LLM APIs, agent frameworks, tool use / function calling, MCP or similar protocols, retrieval, evaluation frameworks (e.g. Langfuse-style), prompt and context engineering, guardrails and safety patterns.
• A pragmatic view of where LLMs actually add value and where they over-complicate things — and the judgement to push back when AI is not the right tool.
• Experience using AI to accelerate engineering itself — e.g. Claude Code, Codex, Copilot, custom skills/agents — and clear opinions on how to make this work at team scale.
• A natural inclination to share what you know: you enjoy explaining things, pairing with others, writing things down and lifting the level around you.
• Independent and action-oriented. You can operate without close supervision, take ownership of ambiguous problems, and move things forward.
• Good communication skills in English across technical and non-technical audiences.

Nice to have:

• Background or interest in measurement, ad tech, martech, analytics, or other data-heavy domains where signals are noisy, biased and probabilistic.
• Experience contributing to or running internal AI enablement / developer productivity programs.
• Depth beyond the API surface — understands why context windows, retrieval and fine-tuning / PEFT behave the way they do, and when each beats the others, not just how to call them.
• A track record of open-source, writing, talks or other public work in the AI / ML space.

Why AudienceProject

At AudienceProject, you will join a team of creative technologists and data scientists redefining how advertising is measured. We combine the rigor of scientific modelling with the pragmatism of product delivery, and we work in close collaboration with major advertising partners such as Google, Meta and Amazon. You will work in an organization that values autonomy, empathy and experimentation, where the impact of your work — both internally and in the systems our global clients rely on — is visible quickly and at scale.

The AI Lab is one of the most strategically important bets we are making right now, with full backing from leadership and direct exposure across the company. If you want to shape how a 100+ person measurement company genuinely becomes AI-native — not just at the prompt level, but in the systems we ship — this is a rare seat to take.

We're reviewing applications and interviewing on a rolling basis, and aiming to have the role closed by mid-August — so if this sounds like you, don't sit on it.

Questions, or want to hear more before applying? Email Martin at [email protected]. Link to the full role in the comments.