Bayer & Microsoft Enhance the ‘Last Mile’ with E.L.Y. AI Model
Editor’s note: In a recent blog on Software is Feeding the World (SFTW), agrifoodtech consultant Rhishi Pethe delves into Bayer’s groundbreaking partnership with Microsoft and the launch of E.L.Y., a small language model (SLM) tailored to improve crop protection efficiency. Here’s a summary of that article:
Rhishi Pethe recently explored Bayer’s groundbreaking partnership with Microsoft and the launch of E.L.Y., a small language model (SLM) tailored to improve crop protection efficiency. Currently benefiting over 1,500 Bayer employees in the U.S., E.L.Y. offers faster, more accurate answers to complex agronomic questions, helping farmers and agribusiness professionals access real-time insights. The model is built on Bayer’s proprietary data, agricultural extension resources, and product catalogs, ensuring reliable, actionable knowledge for stakeholders.
Bayer and Microsoft’s collaboration leverages the Azure platform to provide scalable, customizable, and resource-efficient AI solutions for agriculture. E.L.Y. integrates Retrieval Augmented Generation (RAG), enabling third-party developers to personalize the model to their specific needs. The model’s small architecture makes it ideal for deployment on edge devices, allowing for use in resource-constrained environments.
Rhishi further reflects on the promise of small language models, predicting they will drive a significant shift in the adoption of generative AI at the enterprise level. By targeting agronomists and subject matter experts, this model addresses many challenges in generative AI adoption, especially in legacy industries like agriculture. The approach used by Bayer and Microsoft reduces the hurdles in moving AI projects from concept to production.
MORE BY RHISHI PETHE
Syngenta’s Chief Digital Officer Weighs in on Leveraging AI for Digital Transformation in Legacy Companies
While generative AI projects often struggle to move past pilot stages, the E.L.Y. model has been trained on extensive real-world data, overcoming key challenges like model precision. This has resulted in a 40% improvement in accuracy and time savings of up to four hours per week for frontline employees. Bayer’s proactive approach to keeping the system updated and addressing concerns around policy, liability, and trust has helped overcome barriers to adoption.
By tackling the “last mile” problem, Bayer and Microsoft are setting a new standard for generative AI in agriculture, ensuring practical, trustworthy, and scalable AI solutions that can revolutionize crop protection practices.
For more in-depth coverage, visit SFTW (Software Feeding the World).
Subscribe Today For
