Salesforce Launches Koa
Salesforce and NVIDIA today announced Koa, Salesforce's first CRM reasoning model for Agentforce, built on NVIDIA Nemotron to help agents reason through complex, multi-step workflows and use the right tools to get work done.
Koa was built by post-training NVIDIA Nemotron 3 Super with a proprietary synthetic dataset modeled on enterprise knowledge from nearly three decades of CRM deployments. The result is a reasoning model grounded in business processes, workflows, and operational policies.
Salesforce and NVIDIA are also bringing Nemotron-based models and accelerated computing into Missionforce, extending that same control over model, data, and deployment environment to government and regulated organizations. Together, the companies are bringing mission-specific AI to private clouds, air-gapped networks, and other secure environments.
"The most valuable thing Salesforce has built isn't our platform; it's the accumulated knowledge of how enterprise business actually works. With Koa, the knowledge is put inside the model itself. We trained a reasoning engine that understands the structure of a deal, the lifecycle of a service case, and the workflows that vary across industries. That's a different kind of intelligence, and it runs entirely inside your trust boundary," said Marc Benioff, chairman and CEO of Salesforce, in a statement.
"AI is creating a much larger opportunity for software. Every company needs useful AI, tailored to its knowledge, expertise, and work. NVIDIA Nemotron open models give Salesforce the foundation to turn decades of enterprise expertise into specialized AI with Koa, creating a CRM model that can reason and securely take action," said Jensen Huang, founder and CEO of NVIDIA, in a statement.
The Koa reasoning model and training corpus were built entirely from synthetic scenarios like generating leads, qualifying opportunities, and resolving service cases across 14 industries, including manufacturing, financial services, healthcare, and travel. Each scenario paired a persona with specific tasks, then mapped the sequence of actions and tool calls an agent must take to complete them.
To post-train the model, Salesforce applied Supervised Fine-Tuning (SFT) and reinforcement learning with Group Relative Policy Optimization (GRPO) with NVIDIA NeMo RL, NeMo Gym, and NeMo AutoModel. NVIDIA Nemotron open models gave Salesforce the control to build CRM intelligence into the model itself.