Geospatial Data Scientist

Objective of the role

We’re looking for a versatile and collaborative Geospatial Machine Learning Scientist or Applied Statistician with a deep interest in remote sensing, computer vision and classification problems. You’ll play a core role in developing models, exploring complex datasets, and communicating results that directly influence product and strategy. This position is ideal for someone who enjoys wearing multiple hats and is excited to work in a fast paced, early-stage environment.

Key responsibilities

• Design and implement classification and computer vision models to detect, classify, and segment land cover types and change detection from remote sensing imagery.
• Work with multi-resolution and multi-temporal geospatial imagery (e.g., Sentinel, Landsat, commercial satellites).
• Design and implement machine learning and statistical models that handle spatio-temporal data.
• Analyse datasets (both large and small) from diverse sources (e.g., sensor networks, geospatial APIs, remote sensing).
• Communicate findings clearly and regularly to both technical and non-technical team members.
• Prototype/experiment with algorithms in multiple languages; while Python is ideal, openness to others (e.g., C++) is beneficial.
• Collaborate in small teams.

Skills and Qualifications

Required

• PhD in a quantitative field - such as mathematics/statistics, data science, remote sensing, computer vision, computer science.
• Demonstrated experience with geospatial imagery, computer vision techniques and classification algorithms (e.g., semantic segmentation, object detection, pixel classification, random forest, convolutional neural networks).
• Experience with cloud platforms or geospatial pipelines (e.g., Google Earth Engine).
• Strong programming skills in (preferably) Python.
• Strong statistical background.
• Familiarity with fast and critical review of scientific literature.
• Great communication skills in explaining complex ideas to a range of audiences.
• Willingness to explore unfamiliar tools and languages as needed.

Desirable but not required:

• Experience with geospatial libraries (e.g., GeoPandas, GDAL).
• Demonstrated experience in geostatistical modelling techniques.
• Demonstrated experience working with spatio-temporal datasets and methods (e.g., kriging, spatial autocorrelation, trajectory analysis, or related fields).
• Experience with weather and climate data
• Experience developing time series models
• Familiarity with data pipelines, MLOps, or cloud computing.
• Contributions to open-source or academic publications