AI Scientist

Proxima seeks an AI Scientist to research and develop a pipeline for designing proximity-inducing molecules. The interdisciplinary role focuses on challenging current methods in protein structure prediction and molecular glue design. Proxima’s drug-discovery work combines its Neo-1 all-atom foundation model for structure prediction and molecular generation with structural interactomics data produced using proprietary XLMS technology.

Responsibilities

• Scientifically direct the design and training of large-scale deep learning systems, and develop novel model architectures and training approaches for unsolved scientific problems.
• Collaborate with chemistry, physics and biology experts on feature engineering and cross-disciplinary methods.
• Contribute to top-tier machine learning conferences and journals, and attend core conferences to follow research developments.

Required qualifications

• MS or PhD in Computer Science, Statistics, Applied Mathematics, Computational Biology, Computational Chemistry or a related subject. Candidates with a BS in these areas may also be considered if highly qualified in other requirements or supported by significant work experience.
• At least 4 years of machine learning research experience in industry or academia, strong familiarity with PyTorch, and a record of contributing novel state-of-the-art deep learning methods through industry work or publications.
• Expertise in ideally several topics from the posting, including diffusion and flow-based models, generative image or video models, LLMs, multimodal LLMs, model pre- and post-training, reinforcement learning, distributed training, geometric or equivariant models, and structure-based drug design.
• Ability to connect business problems to models that deliver value and prioritize research accordingly.

Preferred qualifications

Strong machine learning fundamentals; deep NLP or computer vision experience; hands-on ability to design and tune models from scratch; experience scaling large models or developing generative models for molecules or proteins.

The listed locations are Zurich, Boston, New York and Remote. The position is full-time. Listed benefits for US-based full-time employees include equity, a 401(k) with company match, medical, dental and vision insurance, 13 paid holidays, unlimited PTO and sick time, paid parental leave, and in-office lunch; country-specific benefits apply outside the US.