Salesforce Introduces Koa CRM Reasoning Model for Agentforce

Sep 16, 2026
Salesforce introduces Koa, a CRM reasoning model built on NVIDIA Nemotron for handling complex tasks in Agentforce.
Salesforce Introduces Koa CRM Reasoning Model for Agentforce
Image: ,

Salesforce introduces Koa, its first CRM reasoning model for Agentforce, as detailed on the company website. Built on NVIDIA Nemotron open models, Koa is designed to reason through complex, multi step CRM tasks.

Salesforce trained Koa using synthetic scenarios based on 27 years of CRM workflows. The scenarios cover reasoning, tool use, and decisions across more than 14 industries. Salesforce controls the model weights and runs Koa on its own infrastructure, keeping customer data within its systems during training and inference.

Koa is available through the Data Cloud generative models catalog, as an Agentforce model provider across an organization, and within Agentforce Builder for individual agents and subordinate agents. Salesforce says Koa was 11 percent more precise at selecting the correct action in its CRM Bench tests. Customer pilots include 1-800Accountant, Baxter Credit Union, Engine, Formula 1, UChicago Medicine, and Xero.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like Daily AI Brief or Sales & Marketing AI Weekly.

Also, consider following us on social media:

AI Brief AI Brief (X)
Free newsletter

Sales & Marketing AI Weekly

News about AI tools and innovations for Sales and Marketing professionals.

AI Token Webinar

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more

You may also like

Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.