Junior Machine Learning Scientist – Remote Sensing

This role is posted on behalf of LambdAI Space, a startup supported by SGInnovate.

Background

LambdAI Space is a deep-tech company building scalable climate intelligence from satellite data. We work with insurers, financial institutions, governments, and infrastructure operators to detect, quantify, and prevent climate-related risks at scale. Our core technology combines multi-sensor Earth observation data with physics-informed machine-learning models to deliver cost-effective, high-precision insights across large geographies.

The Machine Learning function plays a central role in transforming raw satellite imagery into production-ready models that power our products. The team’s objective is not only to develop accurate models, but to ensure they are robust, testable, and deployable in real-world environments. This includes building end-to-end pipelines covering data preparation, training, validation, deployment, and monitoring.

As a Junior ML Scientist, you will join a small, highly technical team working closely with senior ML scientists, engineers, and product stakeholders. The goal of the team is to move fast while maintaining engineering discipline: shipping models into production, iterating based on feedback, and continuously improving performance and scalability. This role is hands-on and impact-driven, with direct exposure to production systems and real customer use cases.

Job scope

Key Responsibilities

• Implement, train, and validate ML models for optical and radar satellite imagery.
• Build clean, modular training and inference pipelines (PyTorch/TensorFlow).
• Work with engineering to deploy models into production (Docker, CI/CD).
• Support data preparation: tiling, preprocessing, augmentation, metadata cleaning.
• Write tests (unit + integration) and maintain model performance dashboards.
• Collaborate with senior scientists on R&D tasks such as model improvement, feature engineering, and synthetic data workflows.
• Run experiments, document results, and communicate findings clearly.

Minimum Qualifications

• MSc in Machine Learning, Computer Vision, Remote Sensing, Data Science, or related field – OR strong project/portfolio experience.
• Solid programming skills in Python, with experience using PyTorch or TensorFlow.
• Understanding of convolutional models, transformers/attention, or change-detection techniques.
• Some exposure to geospatial workflows (e.g., Rasterio, GDAL, QGIS) or willingness to learn quickly.
• Basic experience using Git, Docker, or cloud environments (AWS/GCP).
• Comfort working with large datasets and debugging data issues.
• Good written and spoken English.

Preferred (Nice-to-Have)

• Hands-on experience with satellite or aerial imagery projects.
• Familiarity with SAR data or multi-modal fusion.
• Experience pushing models to production (CI/CD, containers).
• Basic understanding of ONNX, model optimisation, or GPU workflows.
• Experience with synthetic data generation or augmentation strategies.
• Curiosity for climate resilience, environmental monitoring, insurance, or earth observation.

Interested applicants may apply directly to DTC at: https://central.sginnovate.com/hub/marketplace/openings/bf0b56c8-6c4e-4404-8718-cbf951a2e1d9