Your role

You will be the first hire for our new Polymer AI lab. You will lead the computational chemistry track: DFT screening, machine-learned interatomic potentials (MLIPs), and the bridge between quantum chemistry and our wet-lab dataset, while the polymer synthesis lab comes online in parallel. You report to the CTO and collaborate daily with the Head of Polymers.

Your role is to ensure the computational layer (DFT features, MLIP physics, SCFT, active learning signals) is ready when the experimental data starts flowing. If you do this well, within two years you're the co-architect of a polymer foundation model trained on a dataset that doesn't exist anywhere else in the world.

Your Responsibilities

• Screen monomer candidates via DFT to generate enriched representations (HOMO/LUMO, partial charges, reactivity, polarizability) that topological fingerprints miss
• Evaluate and benchmark public universal MLIPs for fit with our monomer chemistry
• Run MLIP fine-tuning experiments on pilot wet-lab data as it arrives from the synthesis team
• Build the computational data pipeline from DFT features to training-ready representations
• Co-design the active learning loop with the CTO: which candidates should the wet lab synthesize next, based on model uncertainty and DFT priors?
• Set up and maintain the simulation infrastructure on cloud compute
• As the dataset grows: train an in-house MLIP, run MD simulations for virtual screening of block copolymers, and generate synthetic training data at scale
• Manage the compute budget and make the cloud-vs-on-prem recommendation as we scale

Your profile

• PhD and/or 5+ years of experience in computational chemistry, materials science, chemical physics, or a closely related field, with a focus in soft matter physics / polymers
• Hands-on DFT experience (Gaussian, ORCA, VASP, or CP2K) with production-scale calculations
• Working knowledge of machine-learned interatomic potentials (MACE, NequIP, Allegro, SchNet, or equivalent). You've trained and evaluated MLIPs on real systems
• Python fluency and comfort with ML frameworks (PyTorch or JAX), including writing training loops
• Experience connecting computational predictions to experimental validation. You've collaborated with a wet lab and understand what makes a prediction useful in practice
• Published research demonstrating independent scientific contribution

Strong signals:

• Experience with polymer or soft-matter simulations (coarse-grained MD, SCFT, block copolymer phase behavior, polymer thermodynamics)
• Familiarity with active learning or Bayesian optimization for experiment selection
• Experience with molecular simulation engines (LAMMPS, GROMACS, OpenMM, or equivalent)
• Background in concept bottleneck models, physics-informed ML, or interpretable ML for materials
• Contributions to open-source computational chemistry or ML tooling
• Prior work at an industrial R&D lab, where predictions needed to translate into real decisions

What we offer

• Employee Stock Options
• Flexible working hours
• Learning & Development Programme for all our team members
• A once-in-a-career dataset problem: building a foundational dataset for polymers, from scratch
• Direct access to a parallel synthesis platform generating the data you model
• Berlin HQ, European open-source simulation stack
• Gym membership
• Subsidised lunch and impassioned lunch discussions
• 30 days of individual vacation plus up to 5 all-company holidays between the 25th of December and 1st of January
• Regular team events
• Be part of our journey to a future of advanced materials that outperform