Why AI job titles are colliding across Europe, and how WeAreKeen finds the right talent anyway

Money is pouring into AI across Europe, including the Netherlands and Germany. But more investment doesn't just mean more AI jobs - it means more specialized, harder-to-define AI jobs. If you're currently hiring for AI, ML, or data roles, this is the part that actually affects you: why the search is getting harder, why titles are colliding, and how we cut through it.

 

The investment case

European venture funding reached $17.6 billion in Q1 2026, up nearly 30% year-over-year, with roughly half of all European VC funding in 2026 flowing into AI. Locally, that shows up in real rounds:

Globally, the numbers are even larger - Anthropic ($965B valuation, May 2026) and OpenAI ($852B valuation, March 2026), but the more relevant number for our market is that total AI investment in the Netherlands reached roughly €2.8 billion in 2025, and the Netherlands now ranks 4th in the EU AI Index 2026.

 

More investment doesn't just mean more AI jobs, it means more specialized ones. Hiring success increasingly depends on knowing the difference.

 

What it means for hiring

Capital like this turns into headcount fast. A June–July 2026 analysis of AI job postings across eight European countries found roughly 3,000 roles focused on building AI systems for every 450 focused on governing them.

The Netherlands has the highest AI talent density in Europe (10.9 professionals per 10,000 inhabitants) - yet only 21.2% of Dutch AI companies convert to scale-up status, compared with a European average of 31.1%. Globally, 72% of employers report AI-related talent shortages, with demand outpacing supply roughly 3.2 to 1.

Frame 39527 (1)

Why AI hiring is suddenly so confusing

We regularly see companies advertising for an "AI Engineer" when they're actually looking for an MLOps Engineer or an LLM Engineer. The result: weeks of interviews with candidates who were never the right fit.

Client story: One client asked us for an "AI Engineer." After our intake session, it became clear that they actually needed a Machine Learning Engineer focused on production deployment and not someone building LLM applications. That single detail completely changed the candidate pool.

 

We spend enough time inside the business to understand the real problem before searching for a single candidate.

 

What should the AI actually do?

One of the biggest mistakes we see is starting with the technology instead of the problem. Recruiters (and often hiring managers) ask whether they need someone with Python, PyTorch, TensorFlow, or experience building LLMs.

 

Ask what the system should do, not only what tech it should use. The answer points straight at the right role.

 

Here's how we typically map the problem to a role, though titles vary between companies, which is exactly why the intake conversation matters:

  • Predict future outcomes (forecasting demand, spotting customers likely to churn)

    🤖 Data Scientist or Machine Learning Engineer

  • Classify information (detecting fraud, filtering spam, categorizing documents)

    🤖 Machine Learning Engineer or Data Scientist

  • See or hear (reading medical images, detecting objects in video, transcribing speech)

    🤖 Computer Vision Engineer or ML Engineer

  • Understand language (summarizing, translating, extracting information)

    🤖 NLP Engineer or LLM Engineer

  • Generate new content (text, images, code, chatbot responses)

    🤖 LLM Engineer or Generative AI Engineer

  • Perform multi-step tasks autonomously

    🤖 Agentic AI Engineer or LLM Engineer