Building a Scalable Platform for Geospatial AI Model Inference and Fine-Tuning
- Fred Otieno
- Brian Ogolla
- et al.
- 2025
- AGU 2025
Beldine Moturi is a Software Engineer at IBM Research – Africa, Nairobi, Kenya, where she is part of the Accelerated Discovery team – Climate and Sustainability. Her current work focuses on building tools and software for accessing, processing, and onboarding geospatial datasets, fine-tuning pre-trained geospatial foundation models, and running inference on these models. Through this work, she aims to empower developers and researchers to seamlessly derive insights without engineering complexity by abstracting technical processes into intuitive, end-to-end platform experiences.
Previously, Beldine contributed to the SPARK: Biomed Studio project, collaborating with a multidisciplinary team to design a user-friendly, web-based interface that integrates Biomedical Foundation Models (BMFM) for accelerated drug discovery workflows. This work was presented at IJCAI 2024 in the paper SPARK: Harnessing Human-Centered Workflows with Biomedical Foundation Models for Drug Discovery.
Beldine holds a Bachelor’s degree in Economics and Statistics from University of Nairobi, Kenya. She is passionate about advancing AI and software solutions for climate resilience and sustainability and actively engages in the tech ecosystem by contributing to community and knowledge-sharing initiatives, including presenting at PyCon Kenya 2025, mentoring participants during the UNEP Hack4Env 2025 hackathon.