Data Scientist

What we are looking for:

We're looking for an experienced Data Scientist to help shape the future of humanitarian action through CLEAR (Crisis Learning, Early Warning and Anticipatory Action) - NRC's open infrastructure designed to transform fragmented crisis data into timely, actionable intelligence.

CLEAR brings together diverse data sources - including earth observation, conflict data, field reports, and market intelligence - into a shared, open, and federated platform that enables humanitarian teams to anticipate crises and act before they escalate. By closing the gap between insight and action, CLEAR empowers faster, smarter, and more effective humanitarian responses.

As the Data Scientist, you will play a pivotal role in designing and deploying cutting-edge machine learning and deep learning solutions that support early warning systems and humanitarian decision-making. You will develop predictive models capable of identifying emerging crisis patterns, forecasting humanitarian needs, and transforming complex, multimodal datasets into practical insights for field operations.

This is a highly technical, hands-on role suited to someone who enjoys building, training, and fine-tuning AI models from the ground up. You are comfortable working in data-scarce, rapidly evolving environments, proactively sourcing and preparing data, and solving complex challenges where real-world impact matters.

Working closely with the AI Lead, software developers, and data engineers, you will contribute across the full machine learning lifecycle - from model development and optimisation to performance monitoring and deployment. You will also help shape the technical environment in which models are trained and refined, ensuring AI solutions are scalable, robust, and aligned with the operational realities of humanitarian response.

What you will do:

Here are some of your specific responsibilities

• Support the design and implementation of CLEAR’s data architecture.
• Partner with developers and data engineers to design and stand up the environment needed to train and fine-tune models (including data ingestion pipelines, compute and GPU resources, experiment tracking and MLOps tooling) actively shaping that environment rather than waiting for it to be provided.
• Develop machine learning and deep learning models that integrate multiple data streams to detect early indicators of humanitarian crises, combining earth observation and satellite imagery, conflict event data, climate monitoring, economic indicators, and population movement patterns into unified risk assessment frameworks.
• Build computer-vision and remote-sensing models on satellite and aerial imagery (for example flood-extent mapping, building and settlement detection, displacement-site monitoring, infrastructure damage assessment and land-cover change detection) and fuse these earth observation outputs with non-imagery signals.
• Contribute to building automated alert systems that identify emerging crises, calibrating prediction algorithms for different crisis types.
• Create ensemble modelling approaches that combine traditional statistical methods with advanced AI techniques.
• Fine-tune and adapt foundation models and large language models to humanitarian use cases such as document triage, multilingual report analysis and situation summarisation.
• Explore venues for adapting models to evolving crisis conditions through reinforcement learning systems.
• Implement impact-based forecasting systems that translate meteorological, conflict, and economic predictions into specific humanitarian consequences such as displacement volumes, food insecurity levels, and infrastructure damage estimates.
• Build decision trees and recommendation engines that guide field staff through systematic needs assessment processes informed by predictive analytics and historical response data.
• Create automated reporting systems and interactive dashboar