AI-Readiness & Data Automation Postdoctoral Scholar

Lawrence Berkeley National Laboratory; Berkeley (CA); On-site; USD 60,000 - 80,000; Full time

Job summary

Lawrence Berkeley National Laboratory is looking for a postdoctoral researcher to enhance AI readiness of environmental datasets for the U.S. Department of Energy’s ESS-DIVE repository. The role involves developing guidance on AI-ready data, building validation tools, and automating dataset preparation. Candidates should hold a Ph.D. in a related field and possess strong Python programming skills, alongside experience with environmental datasets. The lab emphasizes inclusive values and equal opportunity in its hiring practices.

Qualifications

• Ph.D. in environmental science, earth science, informatics, or a closely related field.
• Strong programming skills, especially Python or comparable scientific programming.
• Experience working with environmental/scientific datasets.

Responsibilities

• Develop guidance for AI-ready data including metadata and formatting requirements.
• Build tools to validate datasets for AI readiness.
• Automate dataset preparation using structured workflows.

Skills

Python programming

Data cleaning

Data analysis

Communication skills

Experience with LLM-assisted workflows

Education

Ph.D. in environmental science or related field

Tools

netCDF

Job description

The Earth and Environmental Sciences Area at Lawrence Berkeley National Laboratory (LBNL) seeks a postdoctoral researcher to develop and curate unique and cutting‑edge AI‑ready data for the U.S. Department of Energy’s ESS‑DIVE repository.

The DOE Biological and Environmental Research (BER) program produces uniquely valuable datasets increasingly used in AI/ML, but many are not AI‑ready due to inconsistent formatting, missing metadata, or incompatible file types.

The selected candidate will join an interdisciplinary team to improve and expand upon how DOE environmental data is prepared for AI to further our understanding of Earth system processes and to enable environmental management. This includes working with ESS‑DIVE users and the broader community to create machine‑readable data products and develop tools and guidance for contributors.

Responsibilities

• Develop practical guidance for what "AI‑ready data" should include that extends beyond the FAIR (Findable, Accessible, Reusable, Interoperable) principles
• Build and extend tools that validate datasets, and check whether they meet AI‑readiness requirements
• Help automate dataset preparation using reporting format templates and structured workflows
• Enable translation of legacy DOE data into AI‑ready formats, lead creation of example AI‑ready benchmark datasets and supporting documentation
• Define AI‑ready data standards: Establish and maintain guidance on metadata and formatting requirements for DOE environmental datasets
• Build automated checks and tools: Develop LLM-supported methods to assess AI readiness and convert datasets into consistent, usable formats
• Drive training and adoption: Create documentation, tutorials, and outreach to promote AI‑ready data practices across the research lifecycle
• Curate benchmark datasets: Select, standardize, and document ESS‑DIVE datasets for AI training and validation
• Support automated workflows: Contribute to developing agent‑based pipelines that streamline data preparation, validation, and integration

Qualifications

• Ph.D. in environmental science, earth science, informatics, or a closely related field
• Experience working with environmental/scientific datasets (cleaning, processing, analysis, synthesis)
• Strong programming skills, especially Python (or comparable scientific programming)
• Experience with LLM‑assisted or agent‑based workflows
• Strong written and oral communication skills, including the ability to explain technical requirements to non‑experts
• Demonstrated record of scholarly or technical contributions (e.g., publications, reports, or significant software contributions)

Desired skills/knowledge

• Experience with metadata standards, data schemas, or FAIR principles, particularly with data formats commonly used in earth/environmental sciences (e.g., netCDF)
• Experience building data pipelines for ingesting and harmonizing data from multiple sources and tracking data provenance
• Familiarity with agentic AI tooling such as Retrieval Augmented Generation (RAG) pipelines, agent skills, and Model Context Protocol (MCP) servers
• Ability/willingness to travel to partner institutions and conferences as needed

Equal Employment Opportunity Employer: The foundation of Berkeley Lab is our Stewardship Values: Team Science, Service, Trust, Innovation, and Respect; and we strive to build community with these shared values and commitments. Berkeley Lab is an Equal Opportunity Employer. We heartily welcome applications from all who could contribute to the Lab's mission of leading scientific discovery, excellence, and professionalism. In support of our rich global community, all qualified applicants will be considered for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, age, protected veteran status, or other protected categories under State and Federal law.

Misconduct Disclosure Requirement: As a condition of employment, the finalist will be required to disclose if they are subject to any final administrative or judicial decisions within the last seven years determining that they committed any misconduct, are currently being investigated for misconduct, left a position during an investigation for alleged misconduct, or have filed an appeal with a previous employer.