Principal Machine Learning Engineer, Structural Biology | Pharma/BioTech | Series A - Drug disc[...]

A leading organization in the drug discovery field is currently looking for a Principal ML Engineer to spearhead the technical direction for their structural biology models. This hands-on, high-impact role offers the opportunity to advance the application of foundational models to complex structural biology challenges. The successful candidate will work closely with the leadership team, serving as the technical authority on machine learning modeling, architecture, and experimentation in this domain. While this role does not involve people management, the candidate will be expected to provide mentorship and guidance to engineers and researchers on technical content.

The ideal candidate will have deep expertise in training and deploying transformer-based models for protein structure prediction and related tasks. Additionally, they should have a strong understanding of how these models are applied within drug discovery workflows. A proven track record in setting strategy, solving complex technical problems, and delivering impactful ML systems is essential.

• Define approaches for data preprocessing, selection, and benchmarking for new training tasks involving protein structures, complexes, and multimodal biological datasets.
• Design and implement extensions to models tailored to specific challenges, such as predicting protein complex interactions and binding affinities, including data processing, benchmarking, and evaluation pipelines.
• Provide mentorship and guidance to team members, assisting with the planning and execution of complex projects related to structural biology modeling.
• Lead the technical strategy for machine learning applications in structural biology, focusing on adapting and expanding foundational models such as those for protein folding and related tasks.
• Influence key decisions regarding model architecture, data infrastructure, and model deployment strategies.
• Work collaboratively with other teams to ensure models address practical needs in scientific discovery.
• Contribute to scientific publications or open-source projects where applicable.
• Develop and maintain scalable, production-ready machine learning systems, including pipelines for training, inference, and deployment.

Expected Milestones:

• By month 3: Take charge of a structural biology modeling project. Create a strategy and experiment plan for adapting foundational models to a key high-value application.
• By month 6: Deliver the initial functional model extension (e.g., binding affinity prediction head), complete with a clear benchmarking framework and a replicable pipeline.
• By month 12: Oversee multiple ML initiatives in structural biology, showcasing significant improvements in model accuracy and practical impact.

You hold a PhD (or equivalent experience) in machine learning, computational biology, or structural biology, with a proven track record of applying machine learning to real-world protein structure or drug discovery challenges. You have extensive experience in building and training transformer-based models (e.g., protein folding models) using frameworks like PyTorch, PyTorch Lightning, or similar. You understand the data challenges in structural biology and are capable of designing scalable preprocessing, training, and evaluation workflows. You have experience delivering machine learning systems at scale, including CI/CD pipelines, model versioning, and distributed GPU-based training. You are proficient with modern MLOps tools and infrastructure, such as Docker, Kubernetes, cloud platforms, and orchestration tools. You are adept at navigating complex technical environments and can deconstruct and execute ambitious modeling initiatives. You understand how structural biology models contribute to the drug discovery process and can align your work with real-world applications.