AI Engineer / Specialist (m/f/d) at Repa Limited

North Holland, North Holland, Netherlands -

Full Time

Start Date

Immediate

Expiry Date

16 Dec, 26

Salary

55000.0

Posted On

17 Sep, 26

Experience

2 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Industry

Information Technology & Services

Description

AI Engineer - Data

This is our first dedicated AI Engineer in the data team: a hands-on, build-and-ship role with unusually broad scope, right at the core of how our product works.

What you'll own

The AI behind our rewards engine. We want to create our next generation of reward design system fully agentically. We want AI to look into the games, generate tailored offers, and evaluate them before they go live, replacing our current inflexible design.
Open-ended AI problems, end to end. Across the data org's verticals, from payer structures to user acquisition and beyond, you'll take the fuzzy "could AI help here?"problems, deeply understand the underlying business logic and turn them into shipped agentic systems, working directly with our Staff Data Scientists, our CTO, and leadership.
Genuinely agentic systems: We want agents, pipelines, retrieval and evaluation, basically a claude code for reward design. You'll decide where AI belongs and, just as importantly, where it doesn't.
The standard for how AI gets built here. As our first dedicated AI Engineer, the patterns, tooling and judgement you set become how AI-fluent data scientists across the team work.

What you'll bring

Own everything, take initiative: you take AI problems from a vague brief to a shipped, working system without being asked — "not my job" isn't in your vocabulary.
Move fast: you start before you're ready, ship the small test, and iterate against real data.
Raise the bar: your standards show up in the systems you ship, not in what you say.
Focus on impact: you know when not to reach for AI, and can name the simpler thing you chose instead to protect what mattered most.
AI-native: agents, LLM pipelines, RAG and modern AI tooling are already part of how you build — and you can explain exactly how you've used them.
You've built real AI systems: you can code (Python), and you've shipped something genuinely agentic — an agent, an LLM pipeline, at minimum a real RAG system — that got used, with solid engineering or applied-ML foundations underneath, not just LLM calls.
Bonus: adtech, marketplaces, recommendations/ranking, or high-scale consumer products.

How To Apply:

Incase you would like to apply to this job directly from the source, please click here

Responsibilities

AI Engineer - Data

This is our first dedicated AI Engineer in the data team: a hands-on, build-and-ship role with unusually broad scope, right at the core of how our product works.

What you'll own

The AI behind our rewards engine. We want to create our next generation of reward design system fully agentically. We want AI to look into the games, generate tailored offers, and evaluate them before they go live, replacing our current inflexible design.
Open-ended AI problems, end to end. Across the data org's verticals, from payer structures to user acquisition and beyond, you'll take the fuzzy "could AI help here?"problems, deeply understand the underlying business logic and turn them into shipped agentic systems, working directly with our Staff Data Scientists, our CTO, and leadership.
Genuinely agentic systems: We want agents, pipelines, retrieval and evaluation, basically a claude code for reward design. You'll decide where AI belongs and, just as importantly, where it doesn't.
The standard for how AI gets built here. As our first dedicated AI Engineer, the patterns, tooling and judgement you set become how AI-fluent data scientists across the team work.

What you'll bring

Own everything, take initiative: you take AI problems from a vague brief to a shipped, working system without being asked — "not my job" isn't in your vocabulary.
Move fast: you start before you're ready, ship the small test, and iterate against real data.
Raise the bar: your standards show up in the systems you ship, not in what you say.
Focus on impact: you know when not to reach for AI, and can name the simpler thing you chose instead to protect what mattered most.
AI-native: agents, LLM pipelines, RAG and modern AI tooling are already part of how you build — and you can explain exactly how you've used them.
You've built real AI systems: you can code (Python), and you've shipped something genuinely agentic — an agent, an LLM pipeline, at minimum a real RAG system — that got used, with solid engineering or applied-ML foundations underneath, not just LLM calls.
Bonus: adtech, marketplaces, recommendations/ranking, or high-scale consumer products.