ServiceNow and NVIDIA Launch 15B-Parameter AI Agent for Enterprises

ServiceNow and NVIDIA have collaborated to launch a new inference model, Apriel Nemotron 15B, designed specifically for handling complex enterprise workflows. The model boasts 1.5 billion parameters and was introduced during ServiceNow's annual customer and partner conference, "Knowledge 2025." Both companies' CEOs, Bill McDermott from ServiceNow and Jensen Huang from NVIDIA, jointly unveiled the model. Apriel Nemotron 15B is developed using NVIDIA NeMo, an AI platform, and ServiceNow's domain-specific data. It is trained on NVIDIA DGX Cloud, hosted by Amazon Web Services (AWS). Despite its relatively modest size compared to trillion-parameter general-purpose models, Apriel Nemotron 15B is optimized for speed, accuracy, and efficiency, making it capable of real-time problem-solving in IT, HR, and customer service within enterprises. The model can handle thousands of workflows simultaneously, enhancing team productivity and response times. AstraZeneca, one of the first companies to demonstrate the practical benefits of Apriel Nemotron 15B, highlighted how AI agents equipped with this model could save employees approximately 90,000 hours annually. This showcases the model's potential to significantly improve operational efficiency and decision-making processes in corporate environments. In addition to the new model, ServiceNow and NVIDIA announced a novel data flywheel architecture that integrates ServiceNow's Workflow Data Fabric with NVIDIA NeMo microservices, including NeMo Customizer and NeMo Evaluator. This closed-loop system leverages workflow data to refine AI responses, ensuring continuous performance improvements. Strict controls and security measures are in place to protect client data, maintaining compliance and privacy standards. The collaboration signifies a shift in enterprise AI strategies from static models to continuously evolving intelligent systems. ServiceNow plans to deploy the AI agents powered by Apriel Nemotron 15B following the conference, marking an essential advancement in their Now LLM service. This deployment aims to push the boundaries of agent-based AI in various industries. Industry experts recognize the significance of this partnership, noting that it paves the way for smarter and more personalized AI solutions. These solutions not only offer faster resolution times and higher productivity but also meet the growing demands for performance and cost efficiency. As ServiceNow continues to lead in enterprise cloud services and work automation, this partnership with NVIDIA underscores its commitment to innovation and customer success. During ServiceNow's "Knowledge 2025" conference, the company expanded its collaboration with NVIDIA, focusing on the development and deployment of intelligent AI agents for business use. A major outcome of this collaboration is the introduction of the Apriel Nemotron 15B inference model, which has been optimized for enterprise workflows through the integration of NVIDIA NeMo and ServiceNow's specialized data. This model is designed to run on NVIDIA's powerful GPU infrastructure, ensuring high performance and low latency. Performance tests reveal that Apriel Nemotron 15B excels in its size category, demonstrating advanced reasoning and decision-making capabilities. It is particularly adept at executing real-time workflows, which are crucial for managing IT, HR, and customer service tasks. The model's compact size and efficient performance make it an attractive solution for enterprises looking to enhance their operations without incurring excessive costs or delays. To further support the model, ServiceNow and NVIDIA revealed plans to integrate selected NVIDIA NeMo microservices into ServiceNow's Workflow Data Fabric. This integration will create a closed-loop data flywheel architecture that continuously collects and contextualizes workflow data, refining the AI model over time. The data flywheel ensures that AI agents can adapt to evolving business needs and deliver increasingly personalized and accurate responses. For instance, an AI agent can quickly resolve billing issues by analyzing past customer interactions and recommending optimal actions, becoming more efficient with each case. Jon Sigler, ServiceNow's Platform and AI Executive Vice President, emphasized the contextual awareness and decision-making abilities of Apriel Nemotron 15B, enabling it to adapt to complex workflows and deliver large-scale personalized experiences. Kari Briski, NVIDIA's Vice President of Enterprise Generative AI Software, noted that the model would serve as the foundation for enterprise-level inference engines, enhancing employee productivity and driving the development of smart digital agents. The Apriel Nemotron 15B model is slated for release in the second quarter of 2025. Beyond the model itself, the companies aim to ensure that large language models (LLMs) and their supported features are smart, measurable, secure, and practically deployable. This joint effort represents a significant milestone in improving the efficiency, precision, and cost-effectiveness of enterprise workflows. Experts in the tech industry believe that this collaboration between ServiceNow and NVIDIA is a game-changer. It promises to deliver more intelligent and personalized AI solutions that address the complexities of modern business environments and economic challenges. As a global leader in enterprise workflow automation, ServiceNow's continued innovation, combined with NVIDIA's renowned computing and AI expertise, positions them to drive transformative changes in how businesses leverage AI technology.