Research Associate, Atmospheric Science, Machine Learning

DeVine Consulting, Inc.•Silver Spring, MD

3d•$90,000 - $110,000•Hybrid

About The Position

DeVine is seeking a full-time Research Associate with experience in Atmospheric Science and Machine Learning (ML) to join our team and support a government customer. This role involves contributing to projects in data modeling, remote sensing, and machine learning, focusing on the scientific analysis of Earth's atmosphere, ocean, and land surfaces. The successful candidate will work on improving weather forecast performance to deliver more accurate weather insights. This position is located in Silver Spring, MD, and remote work may be considered for candidates who meet or exceed all requirements. Only US Citizens will be considered for hire.

Requirements

Graduate degree in atmospheric science, meteorology, computer science, or a related field.
2+ years of experience developing ML models for weather applications.
Strong ML engineering fundamentals, including model training, validation, evaluation, and documentation.
Experience training, running, and verifying AI-based weather prediction models.
Experience working in cloud-based computing environments.
Experience handling large meteorological datasets and common data formats at scale.
Proficiency with modern deep learning frameworks (e.g., PyTorch or TensorFlow).
Familiarity with large geophysical dataset formats (GRIB, NetCDF, ZARR).
Familiarity with cloud-based computing environments (AWS, GCP, Azure).
Strong written and verbal communication skills.
Ability to manage multiple projects and balance competing priorities.

Responsibilities

Conduct innovative research at the intersection of weather prediction and machine learning, including approaches that leverage observations from satellite constellations.
Develop, verify, and document forecast improvements that provide measurable value to customers.
Partner with engineering and product teams to transition research advances into scalable, operational systems.
Communicate results through internal reviews, customer discussions, and, where appropriate, conferences or publications.
Contribute broadly to improving forecasts and overall product performance.