6-Month Traineeship \u007C AI House Amsterdam, Zuidas \u007C Full-time, paid \u007C Intake: 1 October 2026
Most traineeships try to hire a unicorn: someone who codes like an engineer, thinks like a scientist, and pitches like a founder. We know these people evolved over time, and that is where this trainee ship is for. Bring one great personality and together we work on making you the best.
At Prosus AI, we run multiple teams with one objective: shipping AI that reaches 2 billion customers. Getting there takes different kinds of minds. We like all three. But you, as an individual, love one thing. Your job in this application is to tell us which one, and to prove it.
Which one are you?
You get products from 0 to 1. Uncertainty and responsibility are your fuel.
You don't wait for a spec. You'd probably ignore half of it anyway. Hand you an ambiguous problem on Monday and you'll have a working prototype by Friday, held together with strong opinions and clean-enough code. Your side projects have side projects. You've shipped things: a tool, an app, a hackathon build, an open-source repo people actually use.
In the Lab, you'll:
Take applications from whiteboard sketch to production.
AI in the application feels natural, not forced.
Own zero-to-one product builds inside our food delivery, fintech, and marketplace businesses
Build the "queryable company": internal AI systems the whole organization runs on
We'll know you're a Builder if: your GitHub tells a better story than your CV, and the phrase "we need more alignment before we start" makes you physically uncomfortable.
You know how to get from paper to production, and you read papers on the weekend because you want to.
You're the person who saw a preprint on arXiv Tuesday and had a working implementation by Thursday. You don't just follow the field. You interrogate it. You know the difference between a result that's real and a result that's benchmark theater. But here's what separates you from a pure academic: you care whether it ships. A beautiful idea that never touches a user bores you.
In the Lab, you'll:
Turn cutting-edge research into production systems serving billions of requests
Design evaluation pipelines that actually mean something: datasets, automated + human evals, regression control
Push retrieval and memory systems forward: vector search, session memory, long-term memory
Work across NLP, CV, recommenders, and GenAI at a scale most labs can only simulate
Wrestle with LLM inference at scale: latency, throughput, cost
We'll know you're a Researcher if: you have opinions about which recent papers are overhyped, and you can explain why to someone who's never read one.
You see opportunities others miss. You know how to take a technical concept and sell it as a product.
You're technical enough to build, but your superpower is spotting the gap: the user problem nobody's solving, the internal capability that should be a product, the demo that could be a business. You've probably started something (a startup, a side hustle, a student venture), and even if it didn't scale, you learned more from it than from any course. You can walk into a room of engineers and a room of executives and be dangerous in both.
In the Lab, you'll:
Identify and validate new AI product opportunities across a portfolio of 100+ companies
Grow products from first users to real traction: metrics, positioning, iteration
Translate technical capability into business cases that leadership actually funds
Work alongside the founders and startups embedded in AI House Amsterdam, including the Antler residency
We'll know you're an Entrepreneur if: you've built something, and started to think about the customer and story before the first line of code. Bonus if you experienced some cycles of this as well.
Technical skill gets you in the door. Your lane is what differentiates you.
Mastery of one programming language (Python, Go, Rust, C++, or JS) or a strong foundation in ML, statistics, and programming. We don't require both
Understanding of ML fundamentals: supervised/unsupervised learning, neural networks, optimization
Experience manipulating and analyzing data (Pandas, Spark)
MSc, PhD, or equivalent demonstrated experience: self-taught builders with a real portfolio are equally welcome. (Still enrolled? Apply to the Internship, Student Track.)
Bonus: MLOps, cloud platforms (AWS/GCP/Azure), or production ML experience
And across every lane: candour, accountability, curiosity, and a bias to action. People who wait for certainty before they act won't enjoy it here.
- Duration: 6 months, full-time
- Location: AI House Amsterdam, Zuidas (projects may span our global hubs)
- Intake: 1 October 2026 (applications open now)
- Compensation: Competitive salary and benefits
- Visa: Full support for international candidates
This is the most exciting place in the Netherlands to work on AI. Here's the evidence, not the hype:
The scale is absurd. Prosus builds and backs the tech behind everyday life for ~2 billion people: leading positions in food delivery, fintech, and marketplaces across Europe, India, and Latin America. Your work doesn't sit in a demo. It ships.
The money is real. Prosus invests over $100M a year in AI and is openly gunning to build Europe's first trillion-dollar tech company. Ambition isn't a slide here; it's the strategy.
The building is the scene. AI House Amsterdam on the Zuidas is Europe's AI living room: startups in residence, hackathons, the AI Marketplace conference (12,000+ people, the largest AI-in-production event anywhere), meetups, guest lectures, and founders wandering the halls. You don't network after work. Your workplace is the network.
The talent density is unfair. The Netherlands holds 8% of Europe's AI talent with just 2.4% of its population, and Amsterdam concentrates a third of it. You'll work among 750+ Prosus AI specialists across Amsterdam, Berlin, Bangalore, Cape Town, and São Paulo.
Europe needs this to work. The mission is literal: make Europe an AI maker, not an AI taker. You're joining a bet on a continent, not just a program.
Our goal is simple: to have you grow. And we mean it in a way most programs don't:
If your ideas are worth it, we can make it happen together. You have a product concept? Pitch it. A research direction? Argue for it. A better way to run something? Show us. Talent Lab trainees have real ownership from day one, and real ideas get real backing: compute, mentorship, budget, and a path to build it inside a company that funds over 100 of them.
How the growth works:
Real work, not shadowing. You hold a genuine junior role embedded in the AI function, with full accountability for what you ship.
Twice-monthly training days. High-intensity sessions across three pillars: technical depth (advanced AI/ML workshops, production deep dives, guest lectures from leading researchers), professional craft (product thinking, stakeholder management, data storytelling), and career development (influence without authority, navigating ambiguity, your personal leadership style).
A dedicated mentor from day one. One of our senior AI specialists is matched to your lane, guiding your projects and your growth for the full six months.
A path forward. Top performers get priority access to full-time roles across the Prosus ecosystem and a permanent seat in our alumni network.
Our process is built around the three profiles. From your first application, tell us which one you are.
Application: CV plus your lane. Cover letter optional; proof mandatory. Builders: link what you've shipped. Researchers: link what you've read, written, or implemented. Entrepreneurs: link what you've launched, even the failures.
Technical Assessment: a practical take home exercise, scoped to your profile: Builders prototype, Researchers evaluate, Entrepreneurs validate.
The talent is here. The capital is here. The building is literally here: Gustav Mahlerplein, Zuidas, Amsterdam.
The question is: which one are you, and will you seize the moment?
We're committed to building a diverse and inclusive environment and welcome applications from all backgrounds, regardless of race, religion, gender, sexual orientation, age, or disability status.
Prosus: Unlocking an AI-first world \u007C Making Europe an AI maker, not an AI taker
