Accenture recently introduced the Trusted Agent Huddle, which allows the first-of-its-kind multi-system agent collaboration across entire enterprises.
Trusted Agent Huddle includes top tech partners
As part of Accenture’s AI Refinery, the Trusted Agent Huddle will allow secure and seamless agent-to-agent interoperability across partners including Adobe, AWS, Databricks, Google Cloud, Meta, Microsoft, NVIDIA, Oracle, Salesforce, SAP, ServiceNow, Snowflake, and Workday.
The agents developed by enterprise users seamlessly work together in a single platform, allowing organizations to select and manage the right agents for specific tasks and business objectives.
“Collaboration and integration across platforms and systems is key to running AI agents that can reason and complete complex tasks,” said Justin Boitano, vice president of Enterprise AI Software Products at NVIDIA. “Accenture AI Refinery, built with NVIDIA, offers a Trusted Agent Huddle to let enterprises orchestrate interoperable agent teams to drive efficiency and innovation across virtually every business function.”
Accenture AI Refinery is built on NVIDIA AI Enterprise to boost AI across SaaS and cloud AI ecosystems. The new Trusted Agent Huddle can be used with the NVIDIA Agent Intelligence toolkit, which provides seamless, heterogeneous connectivity between agents, tools, and data.
“With the proliferation of AI agents across the enterprise, trust is the only limit to AI reaching its full potential, and seamless and secure multi-system collaboration between agents will further the impact and promise of what agentic AI can achieve,” said Lan Guan, chief AI officer of Accenture. “With the launch of Trusted Agent Huddle, for the first time, companies will be able to connect and orchestrate agents from different enterprise platforms within AI Refinery, enabling boundaryless capability and unprecedented levels of innovation.”
Offering addresses workflows rather than individual processes
Organizations can transform entire workflows, rather than isolated processes within single domains or systems, with Trusted Agent Huddle to publish and onboard agents into AI Refinery through open standardization protocols, such as Agent2Agent and Model Context Protocol.
The Trusted Agent Huddle will also leverage a proprietary algorithm to evaluate and align agent performance. This process enables the future development of an agent trust score. Organizations with pre-existing agents based on cloud-hosted models can seamlessly integrate those agents into AI Refinery without impacting functionality, enabling future agents to be built on the platform.
“With the launch of Trusted Agent Huddle, we’re helping enterprises like FedEx future proof their innovation and AI investments,” said Karthik Narain, group chief executive– Technology and chief technology officer, Accenture. “Collaboration is the competitive differentiation of the future. Companies can best address today’s volatility, while positioning themselves for future innovation and resilience, if they are able to harness AI innovation from across the ecosystem and not be limited to a single agent ecosystem.”
Accenture’s agent builder can be utilized to adapt agents as business requirements change. FedEx is one of those organizations working with Accenture and NVIDIA to use Trusted Agent Huddle to drive greater resilience.
“FedEx is a trusted player at the center of the supply chain ecosystem. We work to make supply chains smarter for everyone, bringing people and businesses together to drive new efficiencies and growth,” said Sriram Krishnasamy, chief transformation officer and chief digital & information officer, FedEx Corporation. “The Trusted Agent Huddle enables agents from different platforms to collaborate as one team and will help FedEx seamlessly engage and orchestrate information across the supply chain to accelerate innovation.”
Accenture and NVIDIA’s recent history of teaming up has led to new services to help customers meet business goals. Read more about how Accenture, NVIDIA, and KION are optimizing supply chains with AI.