Salesforce and NVIDIA unveil Koa CRM reasoning model at Dreamforce 2026

This collaboration signals a shift towards specialized AI solutions in enterprise software, potentially reshaping how businesses manage customer relationships.

• Salesforce and NVIDIA unveiled Koa on September 15, 2026, at Dreamforce, marking a significant collaboration in CRM technology.
• Koa is a specialized reasoning model designed to enhance multi-step workflows in sales, marketing, and customer service, using synthetic data for training.
• Pilot programs are underway with select customers, with general availability expected in Winter 2026 across U.S. regions.

• The partnership builds on previous integrations of NVIDIA's Nemotron models into Salesforce's Agentforce, focusing on domain-specific AI.
• Enterprises are increasingly seeking specialized AI to handle complex workflows without the risks associated with general-purpose models from competitors.
• Koa's training utilized 27 years of synthetic CRM data, ensuring no customer data was used, which addresses privacy concerns in AI applications.

The launch of Koa represents a strategic pivot in the CRM landscape, driven by the growing demand for specialized AI solutions that can handle complex workflows with greater efficiency. Salesforce and NVIDIA's collaboration is not just about introducing a new product; it reflects a broader trend where enterprises are moving away from general-purpose AI models that often come with high token costs and data privacy risks.

Koa is built on NVIDIA's Nemotron 3 Super model, which has been fine-tuned using supervised learning and reinforcement learning techniques. This approach allows Koa to excel in specific tasks such as updating opportunities, routing cases, and scheduling follow-ups, achieving up to three times fewer errors compared to leading general-purpose models. This precision is crucial for businesses that rely on accurate data handling to maintain customer satisfaction and operational efficiency.

The decision to use synthetic data for training Koa is particularly noteworthy. By leveraging 27 years of synthetic enterprise CRM data, Salesforce ensures that Koa can learn from a vast array of scenarios without compromising customer privacy. This is a significant advantage in an era where data sovereignty is paramount, especially for enterprises operating in regulated sectors.

As Koa enters its pilot phase with select customers, including notable names like Formula 1 and UChicago Medicine, the initial feedback has been promising. Early adopters report improved precision in their sales and service workflows, indicating that Koa could redefine how businesses approach customer relationship management. The broader rollout planned for Winter 2026 will likely expand Koa's impact, making it a key player in the CRM market.

This collaboration also highlights a shift in the competitive landscape, as general AI providers may face pressure to enhance their offerings with more domain-specific capabilities. As enterprises increasingly prioritize efficiency and specialization, the demand for tailored AI solutions like Koa is expected to grow, potentially reshaping the future of CRM technology.

• Sales and marketing teams in enterprises will experience enhanced efficiency and accuracy in their workflows.
• Customer service representatives will benefit from improved case routing and follow-up scheduling, leading to better customer experiences.
• IT departments will need to adapt to new AI integrations and ensure data sovereignty is maintained.
• Regulated sectors such as healthcare and finance may see tailored applications of Koa that comply with industry standards.

• Customer feedback from pilot programs will provide insights into Koa's effectiveness and areas for improvement, influencing future iterations.
• Market reactions from general AI providers will indicate how they adapt to the growing demand for specialized models, potentially leading to new offerings.
• Regulatory developments regarding AI in enterprise applications could shape how Koa and similar models are deployed across different sectors.