Position Summary:

We are seeking a motivated Visiting Associate Data Scientist for a part-time (50%), eight-month term to assist our Geospatial AI team. The role involves conducting research and developing advanced generative AI solutions, specifically applying denoising diffusion probabilistic models (DDPMs) to satellite data for forecasting tropical cyclones. This position works under regular supervision and may offer an extension or increase in effort based on performance and funding. The Visiting Associate Scientist may also be involved in the development of proposals for new collaborations involving both academia and industry, as part of joint public-private initiatives aimed at addressing key national challenges.

Essential Duties/Responsibilities:

Satisfy USRA's non-profit mission by making a distinctive impact through scientific and technical accomplishments in the assigned projects.

Train, test and evaluate geospatial machine learning models that utilize satellite data (e.g., denoising diffusion probabilistic models applied to tropical cyclones).

Co-author research papers for leading scientific journals and present findings at relevant scientific conferences.

Engage academic, industrial, and government research communities in collaborative research through activities such as seminars and other talks, workshops, and conferences.

Act as a mentor for student interns involved in machine learning research as needed.

Maintain an awareness of funding opportunities, and support and/or lead proposal activities as approved.

Act safely, ethically and in compliance with all laws and regulations and USRA policies and procedures including:

Ensure the protection of intellectual property including submitting new technology disclosures and supporting patent applications.

Ensure export control compliance related to ITAR/EAR regulations.

Ensure compliance with all policies such as for timesheets and travel.

Complete all compliance trainings as scheduled.

Additional Responsibilities

Model USRA's core values: Discovery, Innovation and Service

Complete ongoing learning and development staying abreast of the state-of-the-art

Other job duties as assigned.

Required Qualifications

Education:

A Master's Degree or higher in, Data Science, Computer Science, Geophysics or related field. PhD preferred.

Experience:

Minimum 2 years hands-on experience training, testing and evaluating generative machine learning models with satellite data applied to terrestrial weather phenomena (experience from employment or internships is applicable)

Experience co-authoring papers and presentations for academic audiences is a plus.

Knowledge, Skills and Abilities:

Strong programming proficiency in Python and extensive experience with core data science tools including PyTorch Lightning, NumPy, XArray, Pandas, Scikit-learn, Lightning, Kubernetes & Git.

Solid theoretical understanding of fundamental machine learning and deep learning concepts (e.g., CNNs, RNNs, transformers).

Excellent problem-solving skills and the ability to work with complex geospatial datasets.

Strong verbal and written communication skills.

Ability to prepare scientific publications as co-author with other scientists

Excellent written and verbal communication in English is required.

Working Conditions:

This is a remote work position.

Travel:

Position may require limited domestic and international business travel.

Universities Space Research Association

Latest Jobs

Vis Assoc Scientist, Data -Geospatial AI