Machine Learning Scientist I/II

Build machine-learning methods that automatically proofread neuron and synapse reconstructions in petabyte-scale optical connectomics volumes, using cell shape and protein barcode information to cut segmentation and proofreading costs. The scientist owns segmentation and synapse-detection pipelines, ingests reconstructions into proofreading tools such as WebKnossos and CAVE, and works with wet-lab scientists and microscopists towards cubic-millimetre reconstructions of mammalian brain tissue.

PhD in computer science, machine learning, applied mathematics, computational biology or a related field (or an equivalent publication record) and 3–5 years of relevant experience; strong computer vision and machine learning on large microscopy datasets; software engineering practice. Graph, sequence or point-based models, chunked cloud data (Zarr on AWS or GCP) and connectomics tools (Neuroglancer, WebKnossos, CAVE, CATMAID) are preferred. Salary $130,000–225,000.