Back

Who we are

With research staff from more than 70 countries, and offices across the globe, IFPRI provides research-based policy solutions to sustainably reduce poverty and end hunger and malnutrition in developing countries.

Danielle Resnick

Danielle Resnick is a Senior Research Fellow in the Markets, Trade, and Institutions Unit and a Non-Resident Fellow in the Global Economy and Development Program at the Brookings Institution. Her research focuses on the political economy of agricultural policy and food systems, governance, and democratization, drawing on extensive fieldwork and policy engagement across Africa and South Asia.

Back

What we do

Since 1975, IFPRI’s research has been informing policies and development programs to improve food security, nutrition, and livelihoods around the world.

Where we work

Back

Where we work

IFPRI currently has more than 480 employees working in over 70 countries with a wide range of local, national, and international partners.

Breaking the Silence: Deploying Automatic Speech Recognition for Agricultural Advisory

Co-organized by IFPRI and CGIAR Program/accelerator Digital Transformation
Webinar Series – AI for Food Systems Research

April 9, 2026

  • 9:30 – 10:30 am (America/New_York)
  • 3:30 – 4:30 pm (Europe/Amsterdam)
  • 7:00 – 8:00 pm (Asia/Kolkata)

Speech-based interfaces are increasingly central to agricultural advisory systems, particularly in low-resource and multilingual contexts where oral communication remains the primary channel for accessing information. Automatic Speech Recognition (ASR) is therefore a critical—but still under-examined—component of AI-enabled agricultural extension. 

This webinar presents emerging research on the state of the art in ASR for agriculture, with a focus on real-world deployment for low-resource languages. It draws on a recent review and benchmarking study examining how current ASR models perform outside laboratory settings, and what it takes to build systems that work reliably in agricultural contexts. 

Nelson Mganga, consultant for IFPRI, will open the session by reviewing existing deployments of ASR in agriculture and key findings from a systematic review of low-resource speech technologies. He will reflect on how these insights shaped IFPRI’s collaboration with Farm Radio International (FRI) to develop and deploy fine-tuned, domain-specific ASR models grounded in farmer-generated audio data. 

A second speaker from Nelson Masuki, Director of Development of GeoPoll will then present results from the technical benchmarking component of the study, comparing the performance of generic and agriculture-adapted ASR models across low-resource languages. The presentation will highlight trade-offs between model accuracy, robustness, and field usability, and discuss implications for applied research and program design. 

The session will be discussed by Joyce Nabende-Nakatumba (Makerere University), who will reflect on the findings from a computer science and language technology perspective, with particular attention to evaluation practices, institutional partnerships, and the future of speech technologies for social impact. 

Together, the webinar aims to advance a more realistic, evidence-based understanding of what ASR can deliver for agricultural advisory systems.

Speakers

  • Nelson Mganga, Consultant, IFPRI; Lead on ASR research and partnership with Farm Radio International
  • Nelson Masuki, Director of Development of GeoPoll  

Discussant

Moderator

  • Eliot Jones-Garcia, Senior Research Analyst, IFPRI; PhD Candidate, Wageningen University