AI-driven geospatial analytics for cost effective crop cutting experiment in Ethiopian Insurance

This Request for Proposal (RFP) represents a collaborative effort between JICA DXLab and an Ethiopian insurance partner to identify a suitable technology solution provider ("Digital Partner") for a Proof of Concept (PoC) in the form of a pilot project. The objective is to develop an AI-driven geospatial analytics solution to optimize Crop Cutting Experiment (CCE) execution, enhance risk modeling for Area Yield Index Insurance (AYII), and improve its administration and scalability. The overarching goal is to scale and deploy the solution for long-term sustainability, supporting smallholder farmers by improving risk assessment and reducing insurance premiums.

This project marks a critical transition from traditional CCE and crop insurance modeling methods to AI-driven approach, integrating geospatial analytics for greater efficiency, cost savings, and improved risk modeling.

Progress reports will document PoC results, activities, and quantitative and qualitative evaluations based on predetermined KPIs. These reports will include technical, operational, and strategic recommendations for JICA and the insurance partner to inform decision-making on future scalability. Should the PoC be successful, JICA may consider extending collaboration with the Digital Partner through a separate contract to refine and expand the implementation of the AI-driven crop insurance solution.

Agriculture is a key sector in Ethiopia, contributing 37% of GDP and employing 70% of the population [1], yet smallholder farmers remain highly vulnerable to climate shocks like drought and erratic rainfall, leading to crop losses, financial instability, and food insecurity. AYII helps mitigate these risks, but the high costs and inefficiencies of manual CCEs have limited its scalability and adoption. The process is inefficient, expensive and resource-intensive, with costs primarily borne by insurance companies and ultimately transferred to farmers through higher premiums, reducing affordability and accessibility.

JICA, in collaboration with a consulting firm ("Operator") and a selected insurance partner, is spearheading efforts to integrate AI-driven geospatial analytics into AYII modeling. This initiative builds on the findings of the Advisor for Index-based Agricultural Insurance Promotion (JICA-AIP), which highlighted the potential of digital solutions in optimizing CCEs and improving risk modeling. By leveraging AI and geospatial analytics, this PoC aims to streamline crop insurance modeling, reduce underwriting costs, and expand accessibility for smallholder farmers.

The long-term goal is to eliminate reliance on manual CCEs, transitioning to a fully digitalized approach for crop yield assessments. By reducing CCE costs by 30–50%, the solution will enhance efficiency, potentially lower insurance premiums, and increase adoption among smallholder farmers. This RFP seeks a Digital Partner to develop and deploy this AI-driven crop insurance solution, ensuring scalability and long-term integration into Ethiopia’s agricultural insurance sector.

[1] ICIP Project Completion Report

We seek a Digital Partner with the necessary infrastructure, proven industry experience, and a track record of delivering AI-driven solutions in agriculture, insurance, or related fields. The partner must comply with Ethiopian data protection regulations, demonstrate financial stability, and be free from corruption.

Additionally, they should propose a digital solution aligned with the PoC objectives, including AI-driven geospatial modeling for index-based crop insurance and a clear, feasible piloting approach with risk analysis and mitigation strategies.

The selection process follows a structured evaluation framework to ensure only qualified candidates proceed. Proposals will be assessed across four key areas:

・Organizational capacity and capabilities - Experience, technical expertise, and prior work in AI-driven solutions.

・Proposed digital solution - AI modeling approach, geospatial integration, scalability, and compliance with project objectives.

・PoC design and support approach - Implementation methodology, pilot execution plan, risk mitigation, and stakeholder engagement.

・Financial proposal - Cost breakdown for each phase, value for money, and feasibility within the project budget.

The selection process will prioritize technical competence and solution effectiveness, ensuring the most qualified bidder is chosen based on their technical proposal and ability to meet project objectives. Shortlisted bidders will be invited for interviews to present and demonstrate their solutions. The highest-ranked bidder following the technical evaluation will proceed to financial evaluation then contract negotiations. This approach guarantees that the selected Digital Partner has the necessary expertise and capacity to deliver a high-quality, scalable solution for AI-driven crop insurance.

The project focuses on developing and implementing an AI-driven geospatial analytics solution to optimize CCE execution, improve risk modeling for AYII, and enhance insurance policy administration. The PoC will be implemented in three phases, with a stage gate after Phase II to determine if additional data collection is needed for model validation:

・Phase I: Pilot Preparation (June – October 2025) - Develop and refine the AI-driven solution, including data collection and landscaping, AI model development for CCE optimization, and capacity building for insurance staff and field teams.

・Phase II: Pilot Execution - Model Deployment (November - December 2025) - Deploy and test the AI model for CCE optimization and yield estimation, conducting the first CCE during the May-June harvest cycle to validate the model’s effectiveness.

・Phase III: Pilot Execution - Model Validation (January – December 2026, Contingent on Phase II Results) - If required, further validate and refine the AI model through a second CCE during the May-June harvest cycle, ensuring robustness across different planting seasons.

At the end of the PoC, a detailed evaluation, reporting, and knowledge transfer process will be conducted. A scalability plan will be developed to explore expanding the solution to additional regions or crops for broader adoption.

The budget for the PoC is capped at USD $410,000, covering all taxes, expenses, insurance, and licensing required for the implementation of the AI-driven crop insurance solution across all phases (Phase I-III). This budget includes costs related to model development, deployment, training, stakeholder engagement, and the execution of both the first and second CCE cycles.

If the PoC proves successful, additional negotiations will take place between the Digital Partner, JICA, and the Operator to allocate further funding for a potential scale-up phase beyond November 2026. This phase would focus on expanding the AI-driven solution to additional regions, integrating new datasets, and enhancing its application for broader index-based insurance products. Payment and compensation terms will be finalized in the contract upon the selection of the Digital Partner.

The project requires AI and geospatial technologies to support index-based crop insurance modeling, including data infrastructure and system integration capabilities for seamless processing, storage, and interoperability with insurance platforms. For this modeling, all necessary data will be sourced by the Digital Partner, except for historical crop yield data, which will be provided by the selected insurance partner, when available.

【Past Event】
We plan to host an information session for prospective bidders on Thursday 29^th May 2025, from 9 -10 AM EAT / 3-4 PM JST. Attendance is optional but requires registration. If you wish to participate kindly fill in this form.

We are happy to address any inquiries from potential bidders via email. Please reach out to us at AI-Crop-Insurance-ET@bcg.com