Intern in the Climate Action, Sustainability and Science Department

Intern in the Climate Action, Sustainability and Science Department - AI for Earth Observation

Job Requisition ID: 21172

Date Posted: 1 November 2026

Closing Date: 30 November 2026 23:59 CET/CEST

Publication Method: External Only

Type of Appointment (learn more):

Directorate: Earth Observation Programmes

Workplace:

Frascati, IT

Location

ESRIN, Frascati, Italy

Our team and mission

You will be part of the ESA Φ-lab. Our mission is to accelerate the future of Earth Observation (EO) by embracing disruptive innovation and being the catalyst for disruptive/transformative innovation in the sector. Our vision is to become an “EO innovation hub” connecting EO with a growing ecosystem of disruptive and transformative innovation like AI, ML, Quantum Computing, Edge Computing, Metamaterials, Photonics, etc. Many of the challenges posed by new technologies still need to be tackled at scientific, application and capability levels to deliver the maximum value from satellite-derived EO assets for our climate, society, and economy. The Φ-lab will bring together early career and senior researchers from a variety of disciplines across EO and disruptive/transformative innovation to contribute to developing innovative EO solutions.

We offer:

• a stimulating multinational, interdisciplinary and open work environment;
• access to high-performance computing infrastructure and unparalleled EO and technology expertise;
• a unique opportunity to work on innovative solutions to address global challenges;
• freedom and focus to conduct creative research while making an impact in relation to ESA’s strategy;
• a wide network of relationships and collaboration with top academia, industry and research centres;
• the opportunity to contribute to the Φ-lab strategy and activities.

Candidates interested are encouraged to visit the ESA website: http://www.esa.int

Field(s) of activity for the internship

Artificial Intelligence (AI) is increasingly transforming Earth Observation (EO) by enabling the extraction of actionable insights from vast and complex datasets. ESA’s Φ-lab is at the forefront of this transformation through its active exploration of cutting‑edge AI4EO technologies, spanning application‑specific AI models, generative models, foundation models, and autonomous AI agents, with applications ranging from ground‑based analytics to onboard satellite payload data processing (i.e., edge computing).

You can choose among the following topics:

• Topic 1: AI for Earth Observation

As part of this internship, you will actively engage with Φ-lab researchers, staff, and Visiting Researchers and support a set of Φ-lab AI4EO research activities, including Geospatial Foundation Models (GFMs) for EO.

You’ll gain hands‑on experience across a number of steps of the AI development cycle: from EO data preparation and model training to benchmarking and testing of developed models.

Tasks may include:

• preparing EO datasets for AI pipelines;
• training and benchmarking AI models for EO applications;
• supporting AI design and optimisation;
• supporting the ideation, preparation, and hosting of Φ-lab Challenges related to AI4EO.

Topic 2: Agentic AI for Earth Observation

As part of this internship, you will develop and evaluate agentic AI for Earth Observation, contributing to research on how agents discover and access EO resources, plan and execute multi-step workflows, interact with external tools, and produce results that are scientifically grounded.

Tasks may include:

• designing agentic workflows that decompose high-level EO objectives into sequences of analysis and processing steps;
• integrating EO tools and data services into agentic systems — data discovery, retrieval, processing, and interaction with specialised models;
• investigating reasoning, planning, and tool‑use strategies that let agents adapt workflows to intermediate results and resource availability;
• building and testing prototype EO assistants, combining language models, geospatial foundation models, scientific models, and external tools as needed;
• benchmarking agentic systems on task completion, tool selection, reliability, efficiency, robustness, and scientific correctness;
• contributing to analysis, technical documentation, and scientific dissemination of results.

You will gain hands‑on experience in agent design, tool integration, scientific workflow automation, and evaluation for geospatial applications — and develop a working understanding of what it takes to make autonomous AI systems reliable, transparent, and scientifically trustworthy in Earth Observation.

Behavioural competencies

• Result Orientation