ServiceNow and Nvidia have partnered to introduce an open-source AI model, Apriel Nemotron 15B, unveiled at ServiceNow’s Knowledge Conference. This collaboration aims to provide cost-effective agentic capabilities for enterprise agents. The new AI model promises lower latency and reduced inference costs through Nvidia’s infrastructure, with access now available.
Apriel Nemotron 15B’s release signifies a major advancement in AI agent services, designed to increase agent availability while cutting costs. An innovative joint data flywheel architecture has been created, combining Nvidia NeMo microservices with ServiceNow’s Workflow Data Fabric. This system enhances AI agent precision and adaptability through a closed-loop learning process.
However, implementing AI agents isn’t without challenges. Many enterprises encounter technical and organizational difficulties, such as tool governance and system vulnerabilities. Almost 40% of businesses find integration with existing systems difficult, according to a Cloudera report. Worker concerns about AI accuracy persist, with some viewing AI output less favorably than human-generated work.
Despite these concerns, optimism persists among CIOs about AI’s potential, especially in IT service desks and data processing. While companies like Walmart have rapidly implemented agent AI, others, including Expedia, remain in the testing phases. Vendors like ServiceNow and Salesforce are addressing apprehensions by developing extensive toolkits to speed up adoption and ease access. Overall, this partnership between ServiceNow and Nvidia marks a significant push to integrate AI agents into enterprise operations, boosting efficiency and reducing costs.