NVIDIA launches the next industrial revolution in knowledge work with open agent development platform

News summary:

• The NVIDIA Agent Toolkit includes the NVIDIA OpenShell open source runtime for building self-evolving agents and claws with improved safety and security.
• The open source NVIDIA AI-Q blueprint for agent search, built on LangChain, ranks high on the DeepResearch Bench accuracy leaderboard. A hybrid approach using both frontier and open models can be used to cut query costs in half.
• Leading software platforms including Adobe, Atlassian, Amdocs, Box, Cadence, Cisco, Cohesity, CrowdStrike, Dassault Systèmes, IQVIA, Red Hat, SAP, Salesforce, Siemens, ServiceNow, and Synopsys are using NVIDIA Agent Toolkit software to advance their enterprise and physical AI agents.
  

GTC—To accelerate a generational shift in software and knowledge work, NVIDIA today announced that it is ushering in the next era of AI with open source software for autonomous, self-evolving enterprise AI agents and working with partners to improve the safety, security, and efficiency of agents.

NVIDIA Agent Toolkit provides an open source model and software for enterprises and developers building tools that enhance productivity by autonomously deciding how to complete assigned tasks. Currently includes: NVIDIA Open Shell™ — An open-source runtime that enforces policy-based security, network, and privacy guardrails that make autonomous agent or claw deployments more secure.

“Claude Code and OpenClaw sparked an inflection point for agents, extending AI across generations and putting inference into action,” said Jensen Huang, founder and CEO of NVIDIA. “Employees will be significantly empowered with a team of state-of-the-art, specialized, custom-built agents to deploy and manage. The enterprise software industry is evolving to specialized agent platforms, and the IT industry is on the brink of its next great expansion.”

NVIDIA Agent Toolkit accelerates community and enterprise software evolution
The NVIDIA Agent Toolkit includes open models such as: NVIDIA Nemotron™, open agents such as NVIDIA AI-Q, and other open skills such as NVIDIA cuOpt™ and open runtimes. open shell. Developers can use Agent Toolkit software to create specialized AI agents that can operate autonomously while using and building on other software to complete tasks.

use of NVIDIA AI-Q Open Agent Blueprints enable developers to create custom AI agents that can recognize, reason about, and act on enterprise knowledge, automatically selecting the right data source and depth of analysis to provide accurate, context-aware answers. A built-in rating system explains how each AI answer is generated.

The AI-Q hybrid architecture uses the Frontier model for orchestration and the NVIDIA Nemotron open model for research. This can reduce query costs by more than 50% while providing world-class accuracy. NVIDIA uses AI-Q Blueprint to deep research bench and Deep Research Bench II Leaderboard.

What NVIDIA OpenShell offers autonomous agent The access you need to be productive while enforcing policy-based security, network, and privacy guardrails. NVIDIA is working with security providers such as Cisco, CrowdStrike, Google, Microsoft Security, and TrendAI to build OpenShell compatibility with cyber and AI security tools.

LangChain, an agent engineering company with an open source framework that has been downloaded over 1 billion times, is collaborating with NVIDIA to integrate its agent toolkit, including AI-Q, OpenShell, and Nemotron open models, datasets, and tools, into the LangChain deep agent library to develop advanced and accurate enterprise AI agents and run them at scale.

Enterprise software platform expands AI agent skills with NVIDIA
Industry software leaders are leveraging NVIDIA Agent Toolkit software, including OpenShell and NVIDIA Nemotron open models, to extend the capabilities of AI agents in their applications and platforms.

• adobe will adopt Adobe’s Agent Toolkit software as the foundation for running hybrid, long-term creative, productivity, and marketing agents in a personalized, more secure, and cost-effective environment.
• Amdocs uses NVIDIA AI-Q and Nemotron to power its Cognitive Core agent platform that continuously monitors customer interactions and billing data to proactively identify and resolve issues before customers are impacted.
• atlassian leverages NVIDIA Agent Toolkit software and NVIDIA OpenShell to continue evolving its Rovo AI agent strategy and AI-powered work systems for tools like Jira and Confluence.
• box uses the NVIDIA Agent Toolkit to enable enterprise agents that use the Box file system to safely and reliably run long-running business processes.
• cadence Leverage Agent Toolkit and Nemotron in conjunction with Cadence ChipStack AI SuperAgent to help engineers design and verify higher quality, more complex semiconductor designs.
• Cisco AI Defense provides AI security protection for OpenShell, adding controls and guardrails to govern agent and claw actions.
• cohesiveness adds support for NVIDIA OpenShell and extends the Gaia AI platform to support more advanced agent workflows with NVIDIA AI-Q.
• cloud strike Announcing Secure-by-Design AI Blueprints that embed Falcon platform protection directly into NVIDIA AI agent architectures, including agents built on NVIDIA AI-Q and OpenShell, and integrates with NVIDIA Nemotron inference models. NeMo Data Designer Power your AI agents to enable investigative workflows.
• Dassault Systèmes is exploring NVIDIA Agent Toolkit software and the NVIDIA Nemotron open model for role-based AI agents called Virtual Companions on the 3DEXPERIENCE agent platform.
• IQVIA integrates NVIDIA Nemotron and other agent toolkit software with IQVIA.ai, an integrated agent AI platform. IQVIA.ai is designed to help life sciences organizations work more efficiently, make better decisions, and responsibly scale AI across clinical, commercial, and real-world operations. IQVIA has deployed more than 150 agents across internal teams and client environments, including 19 of the top 20 pharmaceutical companies.
• Palantir is working with NVIDIA to use Nemotron to develop AI agents running on the Palantir Sovereign AI Operating System Reference Architecture.
• red hat is working with NVIDIA to integrate Agent Toolkit software into Red Hat AI Factory, providing an enterprise-ready platform for building more secure and autonomous agents.
• sales force works with NVIDIA Agent Toolkit software, including NVIDIA Nemotron models, to enable customers to build, customize, and deploy AI agents using Agentforce for service, sales, and marketing tasks. This collaboration introduces a reference architecture that allows employees to use Slack as the primary conversational interface and orchestration layer for Agentforce agents powered by NVIDIA AI infrastructure. Agents can directly participate in business workflows and retrieve data from data stores in both on-premises and cloud environments.
• SAP uses open NVIDIA Agent Toolkit software, including NVIDIA NeMo™, to enable AI agents through Joule Studio on SAP Business Technology Platform, allowing customers and partners to design agents tailored to their unique business needs.
• siemens is launching Hughes EDA AI Agent It uses NVIDIA Nemotron to autonomously coordinate domain-scoped workflows across Siemens’ electronic design automation portfolio for the semiconductor and printed circuit board industries, from design concept to manufacturing approval, improving engineering efficiency and design quality.
• ServiceNow‘s autonomous workforce of AI specialists is built on the ServiceNow AI platform, leveraging NVIDIA Agent Toolkit software, NVIDIA AI-Q Blueprints, and a combination of closed and open models, including NVIDIA Nemotron and ServiceNow Apriel models.
• Synopsys uses NVIDIA Nemotron and Nemo Agent Toolkit to build a multi-agent framework powered by AgentEngineer technology for semiconductor and systems design.
  

Developers can explore NVIDIA Agent Toolkit and OpenShell. build.nvidia.com Currently running on inference providers and NVIDIA cloud partners including Baseten, Bitdeer AI, CoreWeave, DeepInfra, DigitalOcean, GMI Cloud, Fireworks, Lightning, Together AI, and Vultr.

Developers can also use OpenShell with LangChain and download it from: GitHub Run in a dedicated local NVIDIA GeForce RTX™ PCs and Laptops or NVIDIA RTX™-powered workstationsimilarly NVIDIA DGX Station™ and NVIDIA DGX Spark™ from supercomputer altos computing, ASUSDell Technologies, gigabyte, HP, lenovo, MSI and super micro.

Businesses can rely on AI factory infrastructure from cloud service providers such as Amazon Web Services (AWS), Google Cloud, Microsoft Azure, Microsoft Security, Oracle Cloud Infrastructure, as well as Cisco, Dell Technologies, HPELenovo and Super Micro.

Please take a look GTC keynote speech Please explore from yellow session.