Machine Learning Research Engineer

At Flagship Pioneering, we conceive, create, resource, and develop first-in-category life sciences companies to transform human health and sustainability. We’ve created over 100 scientific ventures, including the now familiar drug and vaccine innovator, Moderna Therapeutics.

Since its inception in 2000, many of our companies have leveraging advances in computing, big data and AI. In recent years, this trend has accelerated with first-in-category life science companies such as Generate Biomedicines, Cellarity, Valo and many others that are creating breakthrough innovations using AI and ML technologies.

We are looking for extraordinary computational scientists, engineers, and entrepreneurs to work alongside individuals within the Flagship Ecosystem focused on solving the most impactful challenges in AI and the life sciences.

Position Summary:

We are seeking a machine learning engineer who has experience with the full stack of modern deep learning. This research engineer will focus on developing novel computational kernels and methods to accelerate the development of new ML methods. They will work with our ML scientists to tackle fundamental problems in machine learning (ML), statistics, and information theory, as they apply to Flagship’s work in biology and other domains.  This is an exciting opportunity to be part of a fast-paced, highly dynamic entrepreneurial environment.

Key Responsibilities:

• Identify and mitigate bottlenecks in large neural network training. For example:
• Identifying tasks with significant ring-reduce overhead and developing a faster partial weight-sharing algorithm
• Developing hybrid methods that leverage faster GPU-bound implementations for very large datasets with high memory requirements
• Developing new low-level kernels to increase the expressivity and efficiency of neural networks. For example:
• A new CUDA multiplication kernel that handles a problem-motivated sparsity constraint
• A new PyTorch backpropagation method to sidestep instabilities introduced by autograd.
• Work with the ML team to maintain and extend our cloud research facility.
• Develop clear, intuitive visualizations. Communicate analysis results via presentations to a multi-disciplinary audience
• Cultivate a data-centric and process-oriented company philosophy by helping to maintain best practices for software development, data management, and infrastructure
• Monitor and evaluate new and emerging technologies and models and identify opportunities for collaboration within Flagship Pioneering companies, academia, and third-parties

Basic Requirements:

• MS or equivalent experience in Computer Science, ML, or Software Engineering.
• Fluency in C++ and familiarity with CUDA.
• Fluency in Python and standard ML tools and packages (e.g. Deep Graph Library, PyTorch, Snorkel, etc.)
• Motivated and team oriented, with an ability to thrive in an entrepreneurial and multidisciplinary environment
• Ability to independently lead and run research projects, while maintaining close communication with team members
• Excellent communication and presentation skills. Must be able to speak and ideate with multi-disciplinary team including biologists.  Must be able to think independently, work collaboratively and contribute to an active intellectual environment

Preferred Requirements:

• Experience working with large datasets and training statistical models.
• Familiarity with AWS, GCP, or similar cloud-computing services
• Familiarity with containerization and task orchestration tools (e.g. Docker, Kubernetes, Slurm)