Cassava Technologies Deploys NVIDIA Nemotron Models for AI Factory

Cassava Technologies expanded its AI Factory capabilities by deploying NVIDIA Nemotron open models as a locally hosted, managed inference service. The service is available through the Cassava AI Multi-Model Exchange (CAIMEx), allowing enterprises and service providers to access AI reasoning and tool-use capabilities via APIs without building their own model-serving infrastructure.

To support agentic AI workloads, the company integrated NVIDIA OpenShell for security enforcement and governed agent execution. Cassava also introduced the CAIMEx Token Optimiser, a tool designed to lower computing requirements by reusing previously processed information.

This expansion targets businesses in growth markets across Africa, the Middle East, Latin America, and the United States. By utilizing data centers in South Africa, Egypt, Kenya, Morocco, and Nigeria, the company aims to help organizations deploy AI solutions at scale while maintaining local control and data security.