Analytics and Modeling Intern - Coastal Flood

About the role

Moody's is looking for an Analytics and Modeling Intern for its Coastal Flood Risk team, based in London. The internship lasts 4 months and starts in March 2027. It sits within the Model Development team, which builds mathematical models to assess the financial impact of natural hazards, and focuses on coastal and offshore perils such as storm surge, waves, wave setup and overland inundation.

Key responsibilities

• Support the development and evaluation of machine learning and surrogate models to improve the efficiency and accuracy of coastal flood risk modelling.
• Characterize coastal flooding dynamics, including storm surge, waves and overland inundation, to identify the most relevant predictors and physical processes.
• Enhance existing machine learning models for processes relevant to coastal risk.
• Explore methods such as mSWE-GNN to better represent variability and extreme coastal flooding dynamics.
• Compare machine learning models with alternative surrogate modelling frameworks, including Graph Neural Operators.
• Evaluate model performance in reducing the computational cost of traditional numerical modelling while keeping accuracy high.

Required profile

• Currently pursuing a Master's degree or PhD in Coastal Ocean Modelling, Machine Learning or a related field.
• Ability to complete a 4-month internship starting March 2027.
• Strong analytical skills, with the ability to interpret results and communicate insights.
• Proven ability to implement AI-powered solutions, with a growing awareness of AI risk management and a commitment to responsible and ethical AI use.

Required skills

• High-performance computing clusters in a Linux environment and version control software such as Git.
• Scientific and analytical programming in Python and C++.
• Numerical models for ocean or wave modelling, or statistical modelling approaches.
• Handling and analysing large datasets.
• Familiarity with machine learning methods such as Graph Neural Networks (GNNs) and Fourier Neural Operators (FNOs).
• Model performance evaluation and model inter-comparison methods.