ML Research Scientist (MLRS) - Representation Learning for Molecular AI

The role

You will build rich representations of atomistic systems to develop next-generation foundation models. You will work at the intersection of model architecture, data, and training strategy for microscale world models.

What you will do

• Build a robust pre-, mid-, and post-training curriculum.
• Create reinforcement learning strategies.
• Develop expressive representations of molecular and atomistic structure and dynamics.
• Prototype, benchmark, and iterate research ideas into reusable and scalable components.
• Collaborate with physicists and chemists to ensure models are grounded in real physics.
• Work with research engineers and the infrastructure team to identify support needs.

Required qualifications

• Experience designing, running and analyzing ML experiments at scale.
• Experience with 3D geometric deep learning.
• Demonstrated research impact through conference talks or publications, open-source contributions, or released models.
• Strong interdisciplinary communication and presentation skills.
• Proficiency in Python and modern ML frameworks (PyTorch, JAX).
• Experience collaborating on research projects across multi-person teams.

Nice to have

• Machine learning researcher with professional experience (post-degree) in an industry setting.
• Experience with self-supervised representation learning techniques.
• Experience with Bayesian deep learning techniques and uncertainty quantification.
• Experience developing or applying generative models for 3D systems.
• Experience with equivariant graph neural network architectures (NequIP, MACE, SchNet, PaiNN, or similar).
• Prior experience working in or with researchers in the domains of computational chemistry, biology, or materials science.
• Experience working with multi-cloud distributed compute systems.
• Experience working with multi-site distributed company team.

About the company

Achira is building a team of scientists, ML researchers, and engineers to build a frontier lab for Physical AI for molecules. The goal is to make biology at the molecular level something that can be learned, predicted, and designed.