Senior Machine-Learned Force Field Scientist

Job DescriptionJob Description

About Us

Grafton Biosciences is a San Francisco-based biotech startup focused on solving disease through groundbreaking innovations in early detection and therapeutics. We are combining cutting-edge synthetic biology, machine learning, and manufacturing to fundamentally extend healthy human lifespans. We’re looking for passionate team members who want to shape the future.

Role Summary

As a Senior MLFF Scientist you will lead the end‐to‐end design, training, and deployment of our machine‐learned force‐field models. Your mandate is to achieve sub‐kilocalorie accuracy at GPU scale, steer the quantum‐chemistry and MD data strategy that feeds these models, and integrate the resulting engine seamlessly into pipeline. You will partner with ML, DevOps, and wet‐lab teams to advance our mission.

Key Responsibilities

·       Model development – architect and train neural force‐fields that reach sub‐kilocalorie accuracy on diverse molecular systems.

·       Data pipeline ownership – curate quantum‐chemistry datasets and short MD trajectories; define QC filters and convergence tests.

·       Performance & stability – profile inference on GPUs/CPUs, optimise kernels, and enforce integration stability for long MD runs.

·       Cross‐team integration – expose clean APIs for scoring, gradients and active‐learning hooks; collaborate closely with generative‐model and DevOps teams.

·       Scientific rigor – design validation benchmarks, track regression dashboards, and document model lineage for regulatory audit readiness.

Required Qualifications

• Ph.D. (or equivalent industry experience) in computational chemistry, chemical physics, materials science, or machine-learning for molecular simulation.
• 4 + years hands-on building or deploying machine-learned force fields (e.g., Allegro-/NequIP-style equivariant GNNs, TorchMD-Net, MACE, ANI, SNAP/ACE).
• Proficiency in quantum-chemistry data (DFT and, ideally, higher-level fragments) and molecular-dynamics pipelines (PME, Langevin/RESPA, pyEMMA or similar).
• Strong Python + PyTorch skills and experience profiling/optimising GPU kernels (CUDA, Triton, or comparable).
• Demonstrated ability to deliver models that reach sub-kilocalorie energy/force accuracy on non-trivial chemistries and run at scale (≥105 structures evaluated or ≥100 ns trajectories integrated).

Experience

• Δ-learning over classical force fields or hybrid QM/MM schemes.
• Mixed-precision or INT8 inference optimisation for molecular models.
• Uncertainty-estimation or active-learning loops that decide which QM/MD data to generate next.
• Long-range electrostatics (Ewald/PME, charge equilibration, or neural long-range terms).
• Contributions to open-source molecular-ML or simulation projects (e3nn, OpenMM, LAMMPS plugins, etc.).
• Experience preparing validation packages for regulated or customer-facing environments (GxP, ISO-style, or internal model governance frameworks).

What We Offer

• Competitive salary (DOE).
• Comprehensive health, dental and vision coverage.
• Opportunity to define a new therapeutic‐design paradigm and see your work progress through the clinic.

Company DescriptionGrafton Biosciences is a San Francisco-based biotech startup focused on solving disease through groundbreaking innovations in early detection and therapeutics. We are combining cutting-edge synthetic biology, machine learning, and manufacturing to fundamentally extend healthy human lifespans.Company DescriptionGrafton Biosciences is a San Francisco-based biotech startup focused on solving disease through groundbreaking innovations in early detection and therapeutics. We are combining cutting-edge synthetic biology, machine learning, and manufacturing to fundamentally extend healthy human lifespans.