GeoAI Scientist
FM
FM is a U.S.-based mutual insurance company founded in 1835 that specializes in commercial property insurance for large, highly protected enterprises. It operates on a distinctive engineering-driven model, where risk and premiums are evaluated through site inspections and hazard assessments rather than traditional actuarial methods. With a global presence in over 100 countries and nearly 200 years of expertise, FM emphasizes proactive loss prevention—helping clients identify and mitigate risks before they materialize, earning recognition as one of the top property insurers worldwide
Homepage: https://www.fm.com/
Join our dynamic team at the Science and Technology Center APAC in Singapore, where you will take a leading role in shaping the future of climate risk and resilience across the Asia-Pacific region. As a key contributor within FM APAC Research, you will spearhead geospatial and AI-driven initiatives - guiding the strategic design and deployment of scalable data pipelines, analytical workflows, and intelligent systems that transform spatial and climate data into actionable insights for risk-informed decision-making.
We are seeking a technically exceptional scientist who thrives at the intersection of geospatial data science, artificial intelligence, and climate hazard modeling. You’ll contribute to high-impact research and innovation that translates into intelligent, scalable solutions for real-world risk reduction.
Responsibilities:
- Leverage, curate, and engineer cutting-edge data solutions using geospatial datasets, including geophysical data and climate hazard model outputs - to drive innovation in hazard modeling and spatial impact analysis.
- Lead the evaluation, development, and deployment of advanced AI models.
- Design and automate geospatial workflows using Python and open-source libraries to support scalable analytics.
- Translate research outputs into robust, user-aligned solutions through productization and operational integration.
- Conduct applied research on climate hazards, including tropical cyclones, floods, and extreme rainfall, while collaborating across disciplines to deliver actionable insights and shape the strategic direction of climate risk innovation in APAC and globally.
Qualified candidates must have:
- PhD in Computer Science, Geography, Environmental Engineering, or a related field.
- Fluent in Python and at least one other language (e.g., R, Fortran, Matlab).
- Strong experience in geospatial modelling, particularly in natural hazard contexts.
- Proven ability to implement and operationalize deep learning models.
- Hands-on experience with deep learning frameworks (e.g. PyTorch, TensorFlow, JAX), with demonstrated ability to apply architecture like LSTM, U-Net, and Vision Transformers
- Proficiency with open-source geospatial tools (e.g., GDAL, rasterio, shapely, PostGIS) and GIS platforms
- Skilled in handling large datasets and automating data pipelines via scripting and APIs.
- Excellent communication skills and ability to work in high-performance, research-driven teams.
Desired Skills and Competency Areas
- Experience with HPC environments, GPU parallelization, and cloud platforms (AWS, Azure).
- Familiarity with Databricks for scalable data engineering and machine learning workflows, including Spark-based distributed processing and ML model lifecycle management.
- Familiarity with DevOps practices for scalable model deployment.
- Ability to communicate complex technical insights to non-specialist stakeholders.
- Experience in remote sensing techniques and data analysis, including satellite imagery interpretation, photogrammetry, and LiDAR processing.
- Good understanding of one or more physics-based systems and processes, particularly those related to flood dynamics, hydrometeorological extremes, tropical cyclones, and synoptic-scale weather systems.