The Impact You’ll Make

With treatments for hundreds of diseases in our sights, we’ve built a data science team with domain expertise in machine learning, computer science, bioinformatics, physics, biology, mathematics, applied statistics, and more. We work side-by-side with biologists, automation scientists, chemists, software engineers, and many others; together, we develop the tools and methods to turn our experimental data into treatments for pathologies that affect the lives of countless individuals. As a machine learning engineer, you will design, benchmark, and deploy machine learning models. You will be a part of a multidisciplinary team that will support you in developing your projects from idea to a deployed solution to help Recursion advance drug candidates to the clinic. You will design and build production pipelines that leverage machine learning to make a positive impact for patients.

In this role, you will:

• Build and curate large scale datasets for training machine learning models across multiple modalities
• Develop methods, architectures and benchmarks to train and evaluate models on these datasets
• Productionize these models to integrate them into Recursion OS to improve our drug discovery processes
• Work within dynamic, interdisciplinary teams of biologists, product managers, data and ML scientists, software engineers, and more
• Conceptualize and scope 3-6 month high impact projects for cross-functional workstreams
• Directly contribute to and coordinate execution of these projects
• Mentor other data and machine learning scientists
• Foresee and troubleshoot potential issues in intricate orchestrated systems, and engage appropriate teams for swift and efficient problem resolution.
• Deliver quickly and iteratively, both supporting ad-hoc requests and building improvements for the long-term in agile workstreams

Location:

London | Expected to be on-site 50% of the time.

The Team You’ll Join

Our purpose is to build, test, support, and improve the Machine Learning (ML) models that drive RecursionOS. Our scope of work is vast, including the development of benchmarks, the refinement of our models and maps, and supporting their widespread adoption across the organization.

Our goal is to foster continuous, hypothesis-driven improvements inspired by experimental outcomes. This will be achieved by establishing robust ML operational practices that accelerate model development, testing, and deployment. We will work with all Recursion scientists to develop a clear understanding of our models, uncover new biology & chemistry, make meaningful predictions, and profile the applicability of our methods. Identifying performance gaps and model limitations will inspire new benchmarks, data sources, representations, model architectures, and training strategies.

The Experience You’ll Need

• 3-5+ years of practical experience applying machine learning to solve real-world problems in academic or business settings, with a specific emphasis on leveraging self-supervised and multi-modal learning.
• 2+ years of industry experience in developing, deploying, and maintaining machine learning products in a continuously operational production environment.
• Proven ability to function effectively within a fast-paced, interdisciplinary team, along with the capability to clearly communicate complex theories and techniques to individuals from a variety of technical backgrounds.
• Hands-on experience dealing with at least two types of chemical or biological data, such as microscopy, histology, transcriptomics, proteomics, and molecular structures.
• High proficiency in Python, particularly the data science and machine learning libraries such as PyTorch, NumPy, Sklearn, and Pandas.
• Experience with cloud environments and large-scale distributed systems, specifically GCP, BigQuery, Docker, Kubernetes, and CUDA.
• Demonstrated participation in collaborative software development and peer code review processes, coupled with a solid understanding of Git as a version control tool.
• Proven skills in creating compelling visualizations of high-dimensional data to facilitate clear decision-making and interpretation.

How You’ll be Supported

• A peer trail guide to support you as you onboard and get familiar with Recursion systems.
• Regular 1:1s with your supervisor provides opportunities for support and feedback.
• Receive real-time feedback on code quality and best practices from a team of peers.
• Ability to participate and learn from your colleagues in regular all-hands, journal clubs & tech talks.
• Collaborate with other teams at Recursion, including Valence Labs at Mila in Montreal
• Option to publish research and/or attend conferences to enhance your skills, network with colleagues, and gain insights from a broader professional community.
• Opportunities to participate in open source code development.