NVIDIA Expands Enterprise Agent Toolkit

NVIDIA expanded its open-source Agent Toolkit with tools for enterprises to connect agents to company data, manage their actions and govern their operation. The toolkit combines NVIDIA Agent Knowledge Foundation and CUDA-X libraries with open models, a secure runtime, agent blueprints, and tools such as NeMo Relay and NeMo Compass.

Agent Knowledge Foundation is designed to help agents find relevant company information, interpret it for business use, make predictions and support decisions. Companies can use it without moving data, rebuilding workflows or changing governance controls. Its components include CUDA-X libraries cuDF, cuGraph, cuML, cuOpt and cuVS, plus the new Auto-Ontology library, which organizes raw data into business concepts such as customers, products and orders and maps their relationships. Nemotron Retriever and Kumo Structured models—Kumo Tabular, Kumo Relational, Kumo Time Series and Kumo Forecast—help agents retrieve information and make predictions.

NVIDIA says OpenShell sets a secure boundary on what agents can access and do, while NemoClaw blueprints offer reference designs for specialized agents. A new optimization blueprint provides steps for improving agents over time. In deployments, NVIDIA's supply chain organization uses the foundation through Palantir Foundry, Ontology and AIP. Databricks pairs Kumo Anomaly with Genie Ontology, and Snowflake customers can use Kumo Engine and Kumo Relational in Notebook environments. Amazon OpenSearch Service improved search relevance with fewer steps and virtually no added latency; SAP's fine-tuned reranker ran nearly 3x faster; and ServiceNow improved reranking quality while cutting latency by nearly a quarter. Zoom is using Nemotron Retriever to improve document understanding, search relevance and context retrieval. AWS, ClickHouse, Cloudera, Cohesity, Dassault Systèmes, Dell, Elastic, Exa, IBM, NetApp, Oracle, Perplexity, Starburst and VAST Data are also building on the foundation.

NVIDIA expanded its open-source Agent Toolkit with tools for enterprises to connect agents to company data, manage their actions and govern their operation. The toolkit combines NVIDIA Agent Knowledge Foundation and CUDA-X libraries with open models, a secure runtime, agent blueprints, and tools such as NeMo Relay and NeMo Compass.

Agent Knowledge Foundation is designed to help agents find relevant company information, interpret it for business use, make predictions and support decisions. Companies can use it without moving data, rebuilding workflows or changing governance controls. Its components include CUDA-X libraries cuDF, cuGraph, cuML, cuOpt and cuVS, plus the new Auto-Ontology library, which organizes raw data into business concepts such as customers, products and orders and maps their relationships. Nemotron Retriever and Kumo Structured models—Kumo Tabular, Kumo Relational, Kumo Time Series and Kumo Forecast—help agents retrieve information and make predictions.

NVIDIA says OpenShell sets a secure boundary on what agents can access and do, while NemoClaw blueprints offer reference designs for specialized agents. A new optimization blueprint provides steps for improving agents over time. In deployments, NVIDIA's supply chain organization uses the foundation through Palantir Foundry, Ontology and AIP. Databricks pairs Kumo Anomaly with Genie Ontology, and Snowflake customers can use Kumo Engine and Kumo Relational in Notebook environments. Amazon OpenSearch Service improved search relevance with fewer steps and virtually no added latency; SAP's fine-tuned reranker ran nearly 3x faster; and ServiceNow improved reranking quality while cutting latency by nearly a quarter. Zoom is using Nemotron Retriever to improve document understanding, search relevance and context retrieval. AWS, ClickHouse, Cloudera, Cohesity, Dassault Systèmes, Dell, Elastic, Exa, IBM, NetApp, Oracle, Perplexity, Starburst and VAST Data are also building on the foundation.