🛠 Tech Stack

💼 About This Role

You'll build and lead a new ML Research team within an AI Applications org, shaping its scientific direction while remaining hands-on in model development. Your core impact will be improving co-folding models and advancing their generalization across complex biological datasets for drug discovery. This role combines cutting-edge research with real-world deployment of production-grade AI models.

🎯 What You'll Do

• Establish and lead the ML Research team, defining scientific vision and direction.
• Drive design and training of large-scale foundation models for structural biology.
• Develop data pipelines and model architectures using proprietary datasets.
• Translate research into production-ready modeling approaches for drug discovery.
• Lead hands-on experimentation and model evaluation, especially co-folding models.

📋 Requirements

• PhD or MSc in Computer Science, Machine Learning, or Computational Biology with 7+ years experience including 3+ in technical leadership.
• Strong expertise in applying ML to structural biology (co-folding, protein modeling).
• Proven publication record in top-tier ML or computational biology venues (NeurIPS, ICML, ISMB).
• Hands-on experience with Python and PyTorch and familiarity with large-scale models (OpenFold, Boltz).

✨ Nice to Have

• Experience in early-stage biotech or building ML research functions from scratch.
• Experience with distributed training across GPU/cloud platforms (AWS, Azure, Lambda).
• Familiarity with ML infrastructure and MLOps, including Kubernetes-based workflows.

🎁 Benefits & Perks

• 💰 Competitive compensation including virtual share options.
• 🏠 Remote-first working with flexibility to work from anywhere.
• 🌿 Wellbeing budget including mental health support and home office allowance.
• 📚 Learning budget and generous holiday entitlement.
• 🤝 Regular in-person gatherings at Berlin HQ (approx. 3x/year).

📨 Hiring Process

Estimated timeline: 2-4 weeks · AI estimate

1. 1Recruiter screen· 30 min
2. 2Technical interview· 60 min
3. 3Onsite (virtual) panel· 90 min