• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Salesforce Koa reportedly trained from the same files used to configure agents

    TechCrunch reports that Salesforce’s Koa is built on Nvidia’s open-weight Nemotron model and trained for sales, marketing and customer-support tasks.

    elvisEL
    Silvio SavareseSS
    TechCrunchTE
    5 Sources, ,

    TLDR

    TechCrunch reports that Salesforce’s Koa builds on Nvidia’s open-weight Nemotron model for sales, marketing and customer support.

    A user summarizing Salesforce’s report says the company expands Agent Script files—the specifications that define Agentforce agents—into multi-turn training tasks with simulated user personas. The user cites a Tau2Bench score of 69.41 for Koa versus 68.64 for its base model, and function-call accuracy rising from 0.71 to 0.77.

    Combined views

    275K

    5 Sources, first seen 22d ago

    Combined views

    275K

    5 Sources, first seen 22d ago

    310 likes
    22d ago
    first seen 22d ago
    310 likes
    49 comments
    165 saves
    74 reposts
    49 comments
    165 saves
    74 reposts

    Sentiment

    Positive85.7%14.3%Negative

    Based on 28 sentiment-bearing replies from 28 accounts across 2 conversations.

    Sentiment

    Positive85.7%14.3%Negative

    Based on 28 sentiment-bearing replies from 28 accounts across 2 conversations.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    5 Sources

    TechCrunch@TechCrunchSalesforce Koa is built on Nvidia's open-weight Nemotron model and is trained to do sales, marketing and customer-support tasks. https://spr.ly/6014BGRmwY22d
    elvis@omarsar0Banger report from Salesforce. Pretty interesting to see more of these custom enterprise models. Salesforce trained the enterprise agent model from the same files it uses to configure agents. Koa starts from the open-weight Nemotron-3-Super-120B. Salesforce takes Agent Script specifications, the declarative files that define Agentforce agents, and expands them into multi-turn tasks with simulated user personas. The reward checks whether the agent resolved the task with the right tool calls, and training uses GRPO. The gains are modest and consistent. Koa scores 69.41 on Tau2Bench against 68.64 for its base and 54.48 for GPT-4.1. On CRM Bench it reaches 0.86, close to Claude Opus 4.8 at 0.87, and function-call accuracy rises from 0.71 to 0.77. If your company already describes its workflows in a structured format, those descriptions might be useful to turn into RL environments. Paper: https://arxiv.org/abs/2609.15066 Chat with Paper: https://academy.dair.ai/papers/salesforce-koa-an-enterprise-language-model-for-agentic-tool-use-2609.1506622d
    Silvio Savarese@silviocinguettaWe announced Koa at #DF26 this week. #EnterpriseAI runs on governed, repeatable work, and that takes domain-specific reasoning. Built on @Nvidia's model, @SFResearch trained Koa on 27 years of enterprise workflows. @justjayesh's and my piece for more: https://sforce.co/4rhI9WA20d
    Salesforce AI Research@SFResearchRT @silviocinguetta: We announced Koa at #DF26 this week. #EnterpriseAI runs on governed, repeatable work, and that takes domain-specific r…19d

    5 Sources

    TechCrunch@TechCrunchSalesforce Koa is built on Nvidia's open-weight Nemotron model and is trained to do sales, marketing and customer-support tasks. https://spr.ly/6014BGRmwY22d
    elvis@omarsar0Banger report from Salesforce. Pretty interesting to see more of these custom enterprise models. Salesforce trained the enterprise agent model from the same files it uses to configure agents. Koa starts from the open-weight Nemotron-3-Super-120B. Salesforce takes Agent Script specifications, the declarative files that define Agentforce agents, and expands them into multi-turn tasks with simulated user personas. The reward checks whether the agent resolved the task with the right tool calls, and training uses GRPO. The gains are modest and consistent. Koa scores 69.41 on Tau2Bench against 68.64 for its base and 54.48 for GPT-4.1. On CRM Bench it reaches 0.86, close to Claude Opus 4.8 at 0.87, and function-call accuracy rises from 0.71 to 0.77. If your company already describes its workflows in a structured format, those descriptions might be useful to turn into RL environments. Paper: https://arxiv.org/abs/2609.15066 Chat with Paper: https://academy.dair.ai/papers/salesforce-koa-an-enterprise-language-model-for-agentic-tool-use-2609.1506622d
    Silvio Savarese@silviocinguettaWe announced Koa at #DF26 this week. #EnterpriseAI runs on governed, repeatable work, and that takes domain-specific reasoning. Built on @Nvidia's model, @SFResearch trained Koa on 27 years of enterprise workflows. @justjayesh's and my piece for more: https://sforce.co/4rhI9WA20d
    Salesforce AI Research@SFResearchRT @silviocinguetta: We announced Koa at #DF26 this week. #EnterpriseAI runs on governed, repeatable work, and that takes domain-specific r…19d