accessibility.skipToMainContent
All Positions/Principal AI Scientist (Knowledge & Reasoning)
ResearchHybrid

Principal AI Scientist (Knowledge & Reasoning)

We've built an AI system unlike anything else: 456 specialized experts that discover and crystallize knowledge into binary constraints, eight different reasoning modes, complete explainability. Right now, only one person truly understands the mathematics and logic that make it work.

You need to become the second person who understands all of this. Not just the code (though yes, that too), but the deep mathematical and logical foundations. The Binary Constraint Evolution Networks. The Permuted Agreement Popcount primitive. The eight reasoning modes in Nexus. The theory. The why behind every architectural choice. This isn't a typical research position. You won't spend your days trying to squeeze another 0.5% out of some benchmark. You'll be asking fundamental questions: How do we make AI that discovers knowledge? How do we make reasoning that's both fast and explainable? How do we build intelligence that evolves rather than just trains?
Netherlands or Germany
Full-time
Remote Friendly

Ready to Apply?

Join our mission to advance AI

Competitive salary & equity
Work-life balance & flexibility
Make meaningful impact on AI
Applications reviewed weekly

Questions about this role?

Contact our team

Key Responsibilities

Take ownership of Dweve Loom and its 456 experts. Understand how they discover and crystallize knowledge. Make them smarter.
Own the reasoning system in Dweve Nexus. Eight different modes of reasoning, each suited to different problems. Make them work together seamlessly.
Design experiments to push the boundaries of what our system can do. Some will fail. That's how research works.
Develop new algorithms for constraint discovery, for multi-agent reasoning, for knowledge evolution.
Work closely with our Systems Engineer to co-design algorithms and hardware. The best AI is designed with the hardware in mind.
Analyze the emergent behaviors of the system. What patterns do you see? What do they tell us?
Become the person who understands this system's intelligence as deeply as our CTO does. Then make it more intelligent.

Requirements

PhD-level understanding of machine learning, but your interests go far beyond standard deep learning. You've thought deeply about symbolic AI, evolutionary algorithms, probabilistic reasoning, graph neural networks, or other paradigms.
You're a first-principles thinker. You question assumptions. You're frustrated by the limitations of current AI approaches.
Fluent in English (Dutch is lovely but optional for exceptional candidates).
Comfortable with Python and willing to get your hands dirty with code, but you're primarily a scientist and mathematician.
You care more about understanding intelligence than publishing at NeurIPS. (Though we support publishing too.)
You're driven by questions that might not have answers for years. You're okay with that uncertainty because the questions matter.

What We Offer

Competitive salary plus significant equity. We're bootstrapped, which means founders own the majority and equity actually means something.
Freedom to pursue research that matters to you, not just what's trendy.
Access to a unique AI architecture that nobody else in the world has. Your playground is genuinely novel.
Collaboration with engineers who can turn your ideas into production systems in weeks, not years.
€5,000 annual budget for conferences, research resources, collaboration tools, whatever you need.
Flexible hybrid work, comprehensive health coverage, 30 days vacation, and a team that values deep thinking over quick wins.

Ready to join?

If you're the kind of scientist who gets frustrated reading yet another paper about scaling up transformers, who believes that true intelligence requires more than statistical pattern matching, who thinks symbolic and neural approaches should work together rather than compete, then you might be exactly who we need. You'll have freedom that's rare in research, plus something most researchers never get: a production system running real workloads where you can test your ideas immediately.

About Dweve

We're building the future of AI with binary neural networks that use 60-80% less energy than traditional models. Our European-first approach prioritizes privacy, transparency, and sustainable technology.

Join us in creating AI that actually serves people instead of exploiting them. Your work will have direct impact on making AI more accessible, efficient, and ethical.

Department

Research

Advance the state of AI through advanced research and development

Location & Work Style

Netherlands or Germany
Hybrid & Remote Options
Flexible Hours

Application Process

1
Submit application
2
Initial screening call
3
Technical interview
4
Team fit conversation
5
Offer & onboarding

Apply for this Position

Principal AI Scientist (Knowledge & Reasoning)

Step 1 of 520%

Personal Information

Online Presence (optional)