Deep Learning Tool for Building & Tent Footprint Extraction

MSF Austria’s Earth Observation team is seeking a consultant or firm to develop and pilot an in-house deep learning tool for automated building and tent footprint extraction from very-high-resolution satellite imagery. This project will support faster, independent analysis for operational decision-making, including population estimation, logistics planning, and water/sanitation planning in displacement camps. The solution must run on MSF-controlled infrastructure, be usable by EO team members with minimal coding requirements, and include documentation, validation, and user handover. The consultancy is expected to last approximately 3 months, with a start date of early to mid-October 2026, remote work possible, and at least 3 hours/day overlap with Austrian business hours. Budget ceiling is EUR 45,000.