In 2024 and 2025, Fortell.ai partnered with ADRA Canada to pilot the use of artificial intelligence to improve monitoring and evaluation (M&E) in humanitarian food assistance programming. The pilot spanned four countries: Somalia, Yemen, Syria and Haiti, and explored the use of AI for both data collection and reporting. This case study draws on a summary report prepared by Grant Hillier, International Program Associate Director at ADRA Canada, based on his internal review of the pilot.

Deploying Fortell.ai across diverse humanitarian settings

ADRA Canada conducted a pilot study in partnership with Fortell.ai to evaluate the potential of artificial intelligence (AI) to streamline and strengthen its data collection, analysis, and reporting processes in food assistance programming. ADRA Canada’s food assistance programs are built around a standardised Performance Measurement Framework (PMF). This framework tracks key indicators through baseline and endline assessments, which serve as the foundation for understanding the project’s impact on the food security of recipients over the duration of the response, ensuring accountability and driving learning.

Applications in the field

The pilot tested Fortell’s tools in two ways: AI avatars were deployed for direct data collection, and the Fortell platform was used to automate analysis and report generation.

“In Somalia and Haiti, Fortell.ai avatars were deployed to directly collect data from participants. These AI agents conducted interactive surveys in the local language, which enabled ADRA to test the potential in the future of reducing the need for in- person enumerators. The goal was to test whether avatars could reduce bias, increase reach, and lower costs while maintaining data quality.”

“Syria and Yemen did not utilise avatars for data collection due to concerns from local authorities regarding the use of unfamiliar technologies by NGOs, particularly in sensitive political environments. Human enumerators were used instead, but AI was applied post-collection.”

“In all four countries, regardless of whether avatars were used, implementing partners uploaded their baseline and endline datasets into the Fortell.ai platform. The AI system then conducted standardised analysis and generated draft reports, aligned with ADRA Canada’s reporting template.”

Insights from the pilot

The pilot revealed several important outcomes across efficiency, usability, contextual challenges, and learning.

Efficiency gains were among the clearest benefits. As noted in the internal review, “Report generation time dropped significantly. The actual time varies by implementing partner M&E capacity, and is less important than the fact that it is available within 48 hours after the data is uploaded and is therefore more useful for real-time project management decisions.” This rapid turnaround enabled more adaptive decision-making and improved project responsiveness.

In terms of avatar viability, the pilot found promising results in Somalia and Haiti. “Participants responded positively to AI avatars, requiring little support from the accompanying enumerators, confirming their viability and potential to reduce costs and security risks in regions where in-person visits are logistically challenging.”

The project also supported more reflective and responsive implementation. Thanks to the faster turnaround enabled by AI, teams were able to review findings more quickly and adjust their approach as needed. As the report put it, “Faster turnaround enables more real-time reflection and mid-course corrections.” The system also made it easier to track changes from the baseline, “helping project teams make attribution claims with confidence.”

At the same time, the report emphasised that human insight is still essential. “While the AI analysis flagged data trends, it lacks the contextual knowledge of local staff. AI appears to function best as a supportive tool, enhancing efficiency and refining the process, rather than serving as a substitute for local M&E expertise.”

Final reflections

Collaborating with ADRA Canada on this pilot across four countries gave Fortell an important opportunity to test how a standardised data collection approach could be applied in diverse contexts. By collecting and analysing data on key quantitative indicators such as the Food Consumption Score and the Coping Strategy Index, we were able to observe not only how food security and nutritional vulnerability varied from one setting to another, but also how affected populations adapt to meet their needs. Through this process, we gained valuable insights into both the challenges communities face and the strategies they use to cope with them.

We also placed particular attention on adapting both the accent and physical appearance of the avatars to reflect the local populations. In this pilot, that meant developing a Somali- speaking avatar with a natural, regionally appropriate accent, as well as an avatar in Haitian Creole to ensure more authentic and meaningful engagement with participants in Haiti. These insights have informed ongoing improvements to our platform and will continue to shape how we design inclusive, culturally grounded AI tools for future humanitarian and development work.

We are grateful to ADRA Canada and to Grant Hillier for their trust and collaboration throughout this process, as well as to other key team members who supported the pilot across countries. Special thanks to Dr. Willy Lima, Project Director at ADRA Haiti; Sonja Fraser, Project Manager at ADRA Canada; and Moffat Kiprotich, Country Director at ADRA Somalia, for their guidance and partnership on the ground.