Production ML Engineer - Petabyte-Scale Connectomics

UK-Research-and-Innovation-; Cambridge (MA); On-site; USD 69,605 - 81,036; Full time

Job summary

MRC Laboratory of Molecular Biology in Cambridge seeks a researcher to lead deployment, scaling and maintenance of ML systems for petabyte-scale connectomics data. The role focuses on production pipelines for segmentation, detection and large-scale image analysis, across GPU/CPU clusters and on-prem storage.

The post holder will design infrastructure for training, inference, monitoring and data processing, collaborating with scientists and engineers.

Qualifications

• PhD or equivalent experience in machine learning, data science or a related field.
• Extensive Python experience with PyTorch for research to production pipelines.
• Proven ability to deploy ML workflows at scale on GPU/CPU HPC environments.
• Strong software engineering practices: testing, docs, version control and CI.
• Experience with large-scale image data and bioimage/neuroscience domains is desirable.

Responsibilities

• Deploy, maintain and improve production-scale ML systems for large EM datasets.
• Build robust ML pipelines for training, inference, validation and reprocessing.
• Debug across the full ML stack including data loading, storage, and orchestration.
• Design scalable infrastructure for training, inference and monitoring.
• Collaborate with scientists and engineers to ensure reliable, observable workflows.
• Present findings at seminars and contribute to group discussions and publications.

Skills

Python

PyTorch

Linux

Docker

AWS

HPC

ML Deployment

Education

PhD in CS/engineering or related

Tools

Docker

Singularity/Apptainer

Kubernetes

SLURM

AWS

Job description

MRC Laboratory of Molecular Biology in Cambridge seeks a researcher to lead deployment, scaling and maintenance of ML systems for petabyte-scale connectomics data. The role focuses on production pipelines for segmentation, detection and large-scale image analysis, across GPU/CPU clusters and on-prem storage.

The post holder will design infrastructure for training, inference, monitoring and data processing, collaborating with scientists and engineers.