Research Scientist II, Step 1-4

Job no: 504208

Department:
Earth System Science Center

Work type:
Staff Full-time Exempt

Location:

Alabama

Categories:
Research
- Computer Science, Research
- Environmental

The EarthRISE Project Office, within the Lab for Applied Science (LAS) under the Earth System Science Center (ESSC), is seeking a highly motivated Data Science Research Scientist with applied experience in computer science, geospatial artificial intelligence (Geo-AI), machine learning, cloud computing, and remote sensing/geospatial datasets to join our interdisciplinary science team. The Data Science Research Scientist will play a key role in the team by co-developing Earth-observations-driven data solutions and geospatial artificial intelligence-informed services and supporting workforce development activities in support of the EarthRISE program.

The Data Science Research Scientist will also be responsible for supporting an array of cross-cutting thematic needs in expanding the applied research portfolio of EarthRISE’s Data Science efforts as well as LLM and Geo-AI Integration. A successful candidate will have a strong background in the fields of computer science, remote sensing, agentic workflows, and machine learning algorithms. As part of our diverse, highly dynamic group, you will work side-by-side with a team of exceptional scientists to translate Data Science into actionable insights for decision-makers and stakeholders, while fostering your career growth and development.

• Provide Data Science and technical expertise across a set of thematic areas in support of EarthRISE.
• Work together with the Data Science team as well as the program Chief Scientist and activity managers to provide strategic guidance to EarthRISE office on integration of emerging data science best practices in end-user solutions.
• Support EarthRISE office throughout the project lifecycle (planning, execution, and follow-up) to catalyze the use of advanced data science techniques in support of end-user solutions.
• Support or lead open science work via publications, special journal issues, data products/services, and presenting at scientific conferences.
• Contribute to or lead scientific publications.
• Coordinate and collaborate with broader EarthRISE teams at MSFC and other NASA centers.

• Master’s degree in Computer Science, Engineering, Earth Science, Geosciences, Environmental Science, or other highly quantitative discipline (Bachelor's degree and experience in a specialized area may be substituted for a degree).
• Minimum of 1 year of verifiable, full-time work experience in these disciplines.
• Passion for applied research and translating science into action on the ground
• Experience developing, configuring, and evaluating Large Language Models (LLMs), including open-source and commercial models (e.g., Llama, Mixtral, Gemma, GPT-class models) for scientific reasoning and Earth Observation applications.
• Skilled in the use and development with Python, R, JavaScript, C, SQL, and other programming languages, frameworks, and libraries for research, data analysis, and/or modeling
• Skilled in designing and deploying agentic AI workflows using frameworks such as Lang Graph, CrewAI, Auto Gen, Model Context Protocol (MCP), and Agent-to-Agent (A2A) architectures that integrate geospatial tools, APIs, and scientific data sources.
• Demonstrated experience building Retrieval-Augmented Generation (RAG) systems using vector databases, embedding models, knowledge graphs, and Earth science repositories to improve factual accuracy and scientific traceability.
• Experience developing AI evaluation and benchmarking frameworks for assessing hallucinations, reasoning quality, factual consistency, uncertainty, robustness, reproducibility, and task performance using tools such as Lang Smith, Deep Eval, Ragas, MLflow, or equivalent evaluation platforms.
• Knowledge of Responsible AI, Trustworthy AI, and AI governance principles, including explainability, transparency, provenance, reproducibility, model documentation, risk assessment, and human-in-the-loop validation for scientific decision support.

Skilled in LLMOps and MLOps practices…