below marketmost similar roles pay hereabove market
This role pays less than
97%
of similar roles. Most pay
$151,975–$229,568
— the blue band above.
At the midpoint, this role pays about
$113k
versus about
$191k
for comparable roles.
Based on 240 similar postings.
Employer
About Howard Hughes Medical Institute (HHMI)
Howard Hughes Medical Institute (HHMI) is one of the largest private biomedical research organizations in the world, funding basic research and science education to advance human health and knowledge. Industry: Biomedical Research & Science Education
Howard Hughes Medical Institute (HHMI) currently has
4 open roles
on FindRole.
JOB TITLE: AI Engineer - AI+CryoET
The AI Engineer - AI+CryoET joins the AI+CryoET project within AI@HHMI to develop machine learning methods for mesoscale structural biology. Working within a multi-institutional team, you will build supervised and self-supervised deep learning models for particle detection, localization, and structural analysis in cryo-electron tomography data. Your daily work involves identifying gold nanoparticle probes and nucleosome arrangements in chromatin to improve reconstruction quality and understand cellular macromolecule organization. You will utilize Python, PyTorch, and JAX to design training pipelines that handle missing wedge artifacts and sim-to-real transfer from molecular dynamics simulations. The role requires expertise in 3D volumetric image analysis, inverse problems, and high-throughput processing on Linux HPC clusters to solve complex biological problems regarding chromosome structure and synaptic molecular targets.
What does a AI Engineer earn in Virginia?
Median $246150 from 68 postings across 4 companies.
Develop and evaluate deep learning models for detecting and localizing gold nanoparticles and macromolecular particles in cryoET data.
Identify the arrangement and connectivity of nucleosomes in chromatin to understand chromosome structure.
Develop methods to leverage gold nanoparticle detections to improve tomogram reconstruction and tilt-series alignment.
Design and execute rigorous AI model training and evaluation pipelines for 3D volumetric image data.
Manage sim-to-real transfer from molecular dynamics-derived synthetic training data to experimental datasets.
Identify and guide human annotation efforts for model training and proofreading.
Maintain a well-documented codebase to ensure the reproducibility and extension of research results.
Contribute to scientific publications and present research findings at professional conferences.
What we're looking for
Master's or PhD in Computer Science, Applied Mathematics, Physics, Computational Chemistry, or a related field, or an equivalent combination of education and experience.
3+ years of experience training and evaluating deep learning models, particularly on 3D or volumetric image data.
Strong Python skills and proficiency in PyTorch and/or JAX.
Ability to reason about neural network behavior from first principles, including architectural choices, regularization, and training procedures.
Experience with scalable GPU-based computing environments on Linux HPC clusters and high-throughput processing for large-scale data.
Experience with detection, segmentation, or inverse problems in imaging (preferred).
Experience with cryo-EM/ET data processing, tomographic reconstruction, or related inverse problems in imaging (preferred).
Familiarity with molecular dynamics simulations and/or synthetic data generation for training ML models (preferred).