Are you obsessed with agentic AI? So are we.
Here's the honest pitch: making AI agents do real work in production is becoming the most valuable professional skill of the post-AI era. Most engineers will learn it in 2028, on toy projects. You can learn it now, on live systems.
At Staffyou, AI agents already run real staffing operations in production. They schedule shifts, check compliance and talk to thousands of workers. Our challenge: transform the €90B of manual work in the global staffing industry. We have the data and the customers. We need builders.
The Job
You own agentic projects end-to-end: multi-agent orchestration, loop engineering, harness choices (Claude Code, Codex, or your own loops on the raw APIs), MCP tooling, evals. Take a vague business goal on Monday, demo a working agent on Friday.
- Build and ship agent systems that handle real operations — not demos
- Design how agents plan, verify and coordinate, and prove it works
- Think in business terms: fill rates, margins and compliance drive the design
Who You Are
- Studying Artificial Intelligence, Computer Science, Engineering or related
- Problem solver first: you break messy problems into shippable pieces
- Creative and independent. You prototype instead of asking permission
- Curious about the business model, not just the tech
- Strong Python, with Claude Code or Codex as daily drivers (show us!)
- Available full-time or part-time, minimum 2 days/week for 6+ months
The Deal
🧠 The skill: you graduate knowing how to ship production agents, with direct mentorship from engineering leadership · 💰 €20–35/hour based on impact · 🚀 equity potential for those who stay · 🏢 remote-first, Rotterdam office (Wilhelminakade 81) when you want it · 📚 budget for courses, conferences and compute
How to Apply
Don't send a boring CV. Send your CV or GitHub, your best agentic project, and — in 300 words — what excites you about agentic AI to pavle@staffyou.nl, subject "Agentic AI Student - [Your Name]".
We reply within 72 hours. Then: a technical chat, a paid mini-project (4–8 hours), and a decision within the week.
Ready to build the skill everyone else will be catching up on?
