IFAD enhances food systems and agriculture opportunities in fragile rural communities with help from Azure OpenAI Service

Assessing food systems, populations, and climate 

According to the United Nations (UN), nearly 45 percent of the global population—3.4 billion people—live in rural areas of countries with developing economies. Eighty percent of these people live in extreme poverty. “Most depend on small, family farms for their income and sustenance,” says Thomas Bousios, Director and CIO at IFAD (International Fund for Agricultural Development). “But they are also disproportionately poor. That’s where we try to help.”

IFAD is an international financial institution and specialized agency within the UN committed to eradicating rural poverty and hunger by investing in small farmers and rural entrepreneurs. The organization strives to enhance agricultural productivity, improve rural infrastructure, promote inclusive financial services, and foster resilient communities through investments and partnerships with local individuals and organizations.

“So many factors contribute to the challenges in these rural communities,” says Dina Saleh, Regional Director for Near East, North Africa, Europe, and Central Asia at IFAD. ”We try to make sure that our projects are not just dealing with one specific agenda but integrating several factors.” Saleh adds that most rural people rely on food system activities, including growing, harvesting, and assuring food nutrition, along with packaging, processing, transporting, marketing, and distribution of food. 

Every project IFAD invests in must be reviewed and analyzed first, which calls for collecting diverse data across existing food systems, considering the specific needs of populations, and forecasting what actions can make the most impact. In many countries, climate risks and civil unrest must also be factored. In addition, the organization evaluates ways to best empower women and youth. “Using data and accessing data is the first step to better understand the context in which we’re working,” says Selah. 

Bousios adds, “Previously, we had a lot of different reports and analytics tools being used at different locations around the globe.” It became difficult for the organization to collect, translate, and share information. This led to questions about accuracy and challenges in sharing best practices across regions in a timely manner. It also raised the issue of ensuring equal access to technologies across the regions IFAD serves. “It’s very important to us that any tools or technologies we use are transparent, can be used correctly in a multilateral context, and can address questions like bias to ensure adoption is happening in an equitable way,” says Bousios.

Extracting the right data at the right time

With limited time and resources, finding a solution that supports both seamless configuration management and security management, and the democratization of data, was a priority for the organization. After participating in an AI program presented by the UN in collaboration with Microsoft, the organization became interested in what it could accomplish using Microsoft Azure OpenAI Service. 

“We felt that with artificial intelligence, we could get better insights, improve the targeting of the vulnerable communities we serve, and monitor and adopt the analysis for the best outcomes,” says Alvaro Lario, President at IFAD. “When used in a responsible and transparent manner, artificial intelligence has the potential to increase efficiency and development impact, as well as help us understand data that would otherwise be impossible to understand.”

Katherine Meighan, Associate Vice-President and General Counsel at IFAD, adds, “We firmly believe that AI can have a major impact in helping us as an UN institution significantly work more efficiently, work smarter, and work better. The UN Principles on the Ethical Use of AI provides a guidance and framework for IFAD to ensure that work with AI is informed by ethical considerations through all aspects.”

Using Azure and Azure OpenAI Service, IFAD built a centralized data platform that connects data, dashboards, visualizations, and analytics powered by machine learning and AI. The solution gives IFAD personnel across the globe direct access to data in every region IFAD serves, along with training in analytics and machine learning so that they can be actively involved in creating new tools. 

Omnidata users access information through a simple user interface by providing specific search terms or prompts. Azure AI Search (formerly Azure Cognitive Search) scours the IFAD knowledge base to find relevant reports and publications about a term or prompt and scans content to extract the most relevant sentences and information. When users request a search term–specific summary, the solution pulls together the extracted sentences for each publication and displays an AI-generated summary using Azure OpenAI Service, alongside other critical data that users can access via Power BI. 

In addition, IFAD users can contribute new datasets based on their work that are immediately referenced for other projects in other communities around the world. Omnidata also empowers personnel with self-service training tools to develop skills in AI and machine learning at their own pace and develop tools that support their specific work day-to-day. “So far we’ve gotten tremendous feedback; it’s very transformative for teams to make use of datasets for their own thematic areas and their own specializations,” says Bousios.

Connecting global teams

For IFAD, the Omnidata platform has helped bring the global IFAD community together. Teams can engage across the workforce, develop skills together, collaborate on prototypes, and have equal access to a rich database of knowledge and information that supports their daily work in rural communities. “Our staff have all the data they need at their fingertips, allowing them to make rapid, responsive, and evidence-based decisions which is especially helpful in remote, fragile areas,” says Selah. 

But perhaps the biggest advantage for IFAD is the solution’s ability to quickly aggregate massive amounts of data. “With Azure AI and Power BI, we’re now able to quickly combine a lot of different information sources, for example, multiple data indicators and thousands of documents on agriculture and rural development, to do analysis,” says Bousios. “We can now have a better understanding on the impact of factors such as climate, gender, biodiversity, and get timely information on what needs adjustment to then make it happen in the projects.”

“We’re getting new insights that we didn’t have before,” says Lario. “It’s helping us understand data in ways that would otherwise be impossible and has the power to transform the way we conduct business across the board, allowing us to do much more within our limited resource envelope.

Adds Bousios, “This is the real power of the technology. It has the power to go further and further. I think we’re just scratching the surface.”