Prosper Global connects AI innovation to humanitarian impact with Microsoft Elevate

“The biggest challenge is that crises are no longer happening one at a time,” says Scott Onder, Chief Investment and Innovation Officer at Prosper Global (formerly Mercy Corps).

Across the more than 30 countries where Prosper Global works, conflict, climate shocks, and other disasters often overlap, leaving families little time to recover. As needs grow and resources tighten, its teams must draw on their deep local knowledge while finding time to gather information, assess risks, and coordinate responses while working alongside communities to meet urgent needs and support recovery.

AI can help humanitarian teams find information faster, automate repetitive tasks, and make better use of existing knowledge—giving them more time for the human judgment and relationships at the heart of their work. Prosper Global teams were already developing tools to address challenges they encountered firsthand. The next step was finding a way to bring those innovations to more people across the organization.

“We have amazing teams all over the globe, developing innovative approaches to addressing these challenges,” says Karissa Dunbar, Director of Data Services at Prosper Global. “And the difficult part for us as an organization has been identifying how to take those and scale them.”

Prosper Global already had AI policies and guidelines. What it needed was a more consistent way to identify promising pilots, assess their value and risks, and prepare effective solutions for broader use. Beyond addressing whether a technology worked, it meant addressing questions of governance, privacy, ownership, testing, and staff adoption.

“Up until now, the organization has been doing a lot with AI, and it’s been a bit siloed to date, frankly,” Aaron Eubank, Director of Data Science at Prosper Global. The organization needed practical guidance that reflected the realities of nonprofit work, along with the space to turn that guidance into a plan. That need led three Prosper Global staff members to the inaugural Changemaker Fellowship, part of Microsoft Elevate for Changemakers.

Making innovation work across different contexts

Designed for nonprofit and humanitarian organizations, the fellowship combines applied AI learning, expert guidance, and peer collaboration. Participants work on real challenges within their organizations, building actionable plans to adopt AI responsibly and advance their missions. For Prosper Global, it offered a structured path from individual experiments toward solutions that could support more teams and, ultimately, more communities. “Microsoft Elevate for Changemakers really gives us the space and structure that we need to be able to begin that exploration,” Dunbar says.

The program’s relevance to nonprofit work was just as important. Prosper Global was not looking for a generic AI strategy. It needed an approach suited to a mission-driven organization operating across different countries and contexts, with limited resources and significant responsibilities for the people and information entrusted to it.

Dunbar says the fellowship focused on the practical challenges of adopting AI in humanitarian work, rather than treating it as a theoretical exercise. “The learnings are really focused on AI in our context,” Dunbar says. “What does it mean to be a nonprofit organization, resource-constrained and mission-driven, working to adopt tools and technologies that are new and innovative and cutting-edge and something we’ve never seen before?”

The fellows moved through a connected learning experience that combined live and on-demand training, expert guidance, peer learning, and weekly assignments that built toward an AI adoption plan. Along the way, they examined how to assess use cases, identify stakeholders, establish ownership, address governance and privacy requirements, and prepare teams for change.

For Eubank, the sequence made the work manageable alongside his day-to-day responsibilities. “It was very structured and it really encouraged us to put aside that time,” he says. “The content was engaging and it really felt relevant to our needs around AI.”

From learning to responsible adoption

Participants moved from learning about AI to applying it within their organization through a combination of live and on-demand training, expert guidance, peer learning, and weekly assignments grounded in real-world challenges.

The fellowship also built on Prosper Global’s use and exploration of Microsoft technologies, including Microsoft Power BI, Power Platform, Azure, and Microsoft Copilot. Practical Copilot training helped Prosper Global’s three participants explore how generative AI could support their day-to-day work, while on-demand courses gave them an opportunity to bring their Changemaker Fellowship colleagues into the learning and build their AI knowledge more broadly.

Just as importantly, the learning was tied directly to Prosper Global’s work. Rather than completing hypothetical exercises, participants applied what they learned to actual initiatives and use cases. “Every week felt like it had relevant content and then a really relevant deliverable that required us to identify real people and provide details about our use case and our context,” Eubank says.

The team used responsible AI resources from Microsoft and NetHope to review its practices, identify gaps, and consider how governance could support a more consistent path from pilot to broader adoption. Stakeholder mapping and change management exercises helped the fellows think through who should be involved and how a new tool might affect the people expected to use it.

For Nayid Orozco, AI Solutions and Delivery Manager at Prosper Global, the fellowship offered a way to strengthen the organization’s existing AI guidelines while addressing the human considerations that can determine whether a promising tool succeeds beyond its pilot stage. Drawing on guidance from Microsoft and partner organization NetHope, a nonprofit consortium focused on improving technology and connectivity across humanitarian efforts worldwide, the team identified gaps in its policies, clarified ownership and responsibilities, and considered how new technologies might affect the people expected to use them.

“We always think about the benefits of integrating new technology, but we also need to see what the team will lose,” Orozco says. The fellowship prompted his team to examine concerns such as reduced autonomy or control over data and consider how to address them through clearer communication and more thoughtful adoption planning.

Peer collaboration and building trust across the organization

For Prosper Global, that approach is especially important: AI solutions must earn the trust of staff, protect sensitive information, and support human judgment in work that directly affects vulnerable communities.

The importance of trust, communication, and human oversight extends beyond Prosper Global. Connecting with peers gave the fellows an opportunity to explore how other nonprofits were navigating these same challenges—and learn from the approaches they were taking.

The fellowship connected Prosper Global’s participants with peers at other nonprofits facing similar challenges. Those conversations gave them a chance to compare approaches and learn from people working within the same mission-driven constraints. “Being part of the fellowship and allowing us to work with other peers, organizations that are also facing the same challenges that we are facing, that was really fundamental for us,” Orozco says.

The outcome: A clearer path from pilot to practical use

The fellowship helped Prosper Global bring its separate AI initiatives into a more coordinated approach. One existing tool, VERA: Verified Evidence & Research Assistant, gave the participants a practical opportunity to apply what they had learned.

Developed before the fellowship, VERA is an AI-powered research tool initially designed for Prosper Global’s crisis analysis teams. In places such as Syria, Lebanon, Sudan, and the Democratic Republic of the Congo, analysts draw on United Nations reports, local news, humanitarian data, and other sources to understand rapidly changing conditions. VERA is designed to reduce the time spent gathering that information: analysts can enter a research question through a conversational interface, and the tool searches relevant sources and assembles a report to support their work.

By taking on repetitive research tasks, VERA could give analysts more time to interpret findings, collaborate with colleagues, and apply their expertise to decisions affecting communities.

VERA had reached the proof-of-concept stage when Eubank brought it into the fellowship as his practical use case. He explored what it would take to develop the tool further and make it useful to more teams, including user testing, stakeholder engagement, organizational support, and responsible adoption planning. The work also raised broader questions: Which tools have potential beyond their original use? What must be in place before other teams adopt them? And how can Prosper Global expand their use while accounting for different local needs and workflows?

By the end of the fellowship, participants had a more structured approach to answering those questions. “We now have clear AI guidelines and an adoption plan that can support us in, again, taking those pilots into scale-up solutions,” Orozco says.

VERA still requires further development and testing, but for Eubank, the goal remains firmly focused on the people behind the work: “Tools like VERA and other AI solutions for Prosper Global will really free up time for people within the organization to do that more mission-driven, impactful work.”

More time, stronger decisions, greater reach

Beyond VERA, Prosper Global is exploring how AI, analytics, and other technologies could help its teams anticipate risks, use resources more effectively, and support communities through crises. Potential applications include combining climate and weather data to identify emerging threats and using satellite imagery and environmental information to help communities prepare for drought. In each case, local expertise remains essential: technology must provide relevant information and support the judgment of the people using it.

The fellowship has given Prosper Global a more structured way to pursue these opportunities. Its new AI adoption plan can help teams evaluate promising ideas, engage the people who will use them, and determine where broader adoption could make a meaningful difference. An innovation developed in one location could then benefit colleagues elsewhere, with the support and adaptations needed to meet local needs.

“Microsoft Elevate for Changemakers really gives us the combined elements of structure, expertise, and dedicated space and time to think through and ideate for our organization: How do we want to take this forward?” says Dunbar.

The goal is not simply to introduce more technology. It is to reduce duplicated effort, give humanitarian professionals more time to apply their expertise, and strengthen the support they provide to communities.

The new Elevate for Changemakers platform extends that approach for Prosper Global by bringing learning, events, community, and recognition together in one connected experience. Participants can access role-based live and on-demand training, connect with peers and experts, and continue building the skills and organizational capacity needed to move from AI exploration toward responsible adoption.

“Our north star is figuring out how solutions can help strengthen the resilience of communities around the world,” Onder says. For Prosper Global, that is the promise of responsible AI adoption: helping more teams put their knowledge, experience, and care to work where they matter most—with the people and communities they serve.

Discover more about Prosper Global on Facebook, Instagram, LinkedIn, X/Twitter, and YouTube.