The Mindspan Institute is a new non-profit research institute in Cambridge, MA, aimed at building novel tools to analyze the human brain, and scaling them to study its molecules and wiring. We interpret how they contribute to brain functions and dysfunctions, and share our tools and discoveries as freely as possible, to advance human health.

The role

This is an engineering-oriented science role, owning the computational pipeline that turns terabyte-scale image volumes into reconstructions worth simulating. The job is closer to running an industrial machine learning program than to publishing on new architectures: build the benchmark, measure honestly, find where the models break, get more of the right ground truth, retrain, and ship the improvement into production, on a cadence measured in weeks rather than years.

You report to the Head of Computation and Analysis, and work, as needed, with other Mindspan teams (Microscopy and Optics, Expansion Microscopy, Scalable Chemistry, and Automation and Industrialization), using logic and evidence to advance your work so as to boost the performance of all.

We are ready to start work on such pipelines. There are volumes on disk, a pipeline that runs end to end, and a set of known failure modes waiting for someone to attack in priority order. The position is intended for someone who wants automated reconstruction to become fast and trustworthy enough that biology comes out the other end, not only to run existing models on new data. Given the multidisciplinary nature of the job, it is ideal for a person who enjoys problem solving across intellectual boundaries, at the level of concepts as well as of hands-on debugging. It also requires systematic documentation, analysis, and iterative problem solving, to arrive at working pipelines, apply them, and share them.

What you’ll do

         Own and improve the segmentation pipeline end to end: alignment and preprocessing, dense prediction, agglomeration, and the handoff to proofreading.

         Build the benchmark: curate and version ground-truth volumes, define splits that reflect real sample and imaging variability, and define metrics that track downstream cost, not just voxel accuracy (split and merge rates, error-free path length, proofreading hours per millimeter of cable).

         Make evaluation infrastructure: automated, reproducible, and versioned, so every model and pipeline change is scored against the same suite before it reaches production, with regression gates that block quiet degradation.

         Run the data flywheel: mine hard examples from production runs, turn them into targeted annotation requests, fold corrections back into training sets, and retrain on a predictable cadence.

         Close the loop with annotators and proofreaders: build the tooling and quality control that make their corrections usable as training data by default.

         Report the numbers that matter to the institute: throughput, cost per unit volume, and error rates broken out by tissue, staining, and imaging condition, and tell the wet-lab and optics teams which upstream parameters actually move reconstruction quality.

         Documentation and training: keep pipeline and experiment records that others can execute, with the reasoning behind each parameter.

Required qualifications

• Ph.D. in computer science, electrical engineering, applied mathematics, computational biology or neuroscience, or a related field, or equivalent industry experience.
• Demonstrated expertise training and deploying deep learning models for 3D or large-image segmentation on real, messy data.
• Track record of driving measurable model improvement across many iterations, with the benchmark and evaluation work that made the improvement legible.
• Strong software engineering: Python, PyTorch or equivalent, version control, testing, and comfort owning code that other people depend on.
• Direct hands-on experience processing data too large for one machine: chunked array formats, out-of-core and distributed processing, and workflow orchestration on a cluster or in the cloud.
• Excellent engineering judgment and scientific rigor, including the discipline to tell a real improvement from noise.
• Ability to work directly with experimentalists and annotators, and to translate between sample quality, image quality, and model behavior.

Preferred qualifications

• Direct hands-on experience with connectomics reconstruction: affinity or embedding prediction, watershed and agglomeration, flood-filling, skeletonization, or synapse and structure detection.
• Experience with proofreading and annotation platforms (Neuroglancer, CAVE, webKnossos, VAST, or similar) and human-in-the-loop pipelines.
• Experience running a production data engine or active learning loop in industry, in autonomous driving, medical imaging, or a comparable setting.
• Experience adapting large pretrained segmentation or vision models to domain-specific microscopy.
• Experience with registration and alignment of large serial-section or tiled volumes.
• Contributions to open-source scientific software, public benchmarks, or community challenges.

Culture we’re building

The disorders of the brain affect over 1,000,000,000 people around the world: Alzheimer’s, Parkinson’s, stroke, and other conditions. We operate with a sense of urgency, even as we look to understand the brain at its most fundamental levels.

We are hypercollaborative, and focused on impact over short-term markers of productivity. Creativity and failure are the stepping stones to scale, and working together with “strong opinions, weakly held” is our style.

We solve real engineering problems the best way we can: inventing fast when we have to, and using good existing solutions where they exist. We believe hard problems only yield to many disciplines working together, so collaboration is not optional.

We expect people to explore the unknown, which means making mistakes daily, owning them, and learning from them to improve and excel.

We are recruiting a team of scientists and engineers who exemplify all these values.

What we’re looking for

The ideal candidate is a scientist-engineer who naturally crosses disciplinary boundaries. You are intellectually curious, relentlessly practical, and happiest when solving problems that have never been solved before. You would rather ship a pipeline that measurably improves every month than chase a benchmark number that no one downstream can feel.

You have a strong detail orientation and exceptional organizational, planning, interpersonal, and communication skills. You are good at writing, can work independently and lead and manage, and can prioritize and multitask as needed. Self-motivation, commitment to high quality standards, and a service-oriented mindset are all key. You have good judgment, balancing flexibility and focus, action and planning, speed and thoughtfulness, as appropriate. You are open to giving and receiving feedback, constructively and thoughtfully, to advance scientific missions and personal growth.

This is an in-person job. Flexible hours may be required at times.

What we offer

The chance to build one of the core scientific pipelines of a new research institute alongside some of the most adventurous scientists in the field, with real autonomy and a direct line to the people setting the scientific agenda. The institute is organized around end-to-end runs that carry tissue from intake to a reconstruction worth simulating, and your pipeline is where those runs either yield biology or yield nothing. You will get to contribute to addressing one of the most important unmet medical needs of our time, as part of the initial team. You will learn lots of things that probably cannot be learned any other way, as you solve problems, build, and work with others. As Mindspan grows, we expect this role to grow with it. We are excited to invest in someone who wants to take on increasing responsibility over time.

Comprehensive health, dental, vision, and family benefits, and a matching retirement contribution.

Competitive salary ($120,000-159,000), depending on experience.

Apply for this role →

Questions about the role go to jobs@mindspan.org.

The Mindspan Institute is an Equal Opportunity Employer. We are committed to a collaborative and scientifically rigorous environment for everyone who works here.