Salesforce Koa improves enterprise AI tool use with reinforcement learning
Salesforce Koa: An Enterprise Language Model for Agentic Tool Use
Computation and LanguageArtificial IntelligenceMachine Learning
Summary
Using AI to help with complicated tasks often means teaching the AI to use tools correctly in many steps. The authors created Salesforce Koa by training a big language model to perform better at using tools in step-by-step tasks without relying on private data. They designed a method to simulate detailed workflows and reward the AI when it completes tasks successfully. This method helped Salesforce Koa perform better than similar models on tests involving business and public tools.
What this means in practice
- •For enterprise ai developers: Train enterprise AI assistants to better execute multi-step tasks that involve using various business tools based on declarative workflow definitions.
- •For software integration teams: Improve automated workflows in CRM systems by enabling AI agents to interact effectively with tools through specification-driven training.
Authors
Zixiang Chen, Sufeng Niu, Yingchi Liu, Wenting Zhao, Akshara Prabhakar, Shubham Mehrotra, Bin Bi, Zhujun Lan, Katherine Tan, Mohammad Ramezanali, Tulika Manoj Awalgaonkar, Monojit Banerjee, Jielin Qiu, Shiva Kumar Pentyala, Zhepeng Cen, Anupam Tripathi, Ali Ziaei, Regunathan Radhakrishnan, Darvish Lee Shadravan, Shelby Heinecke, Sitaram Asur, Silvio Savarese, James Zhu, Phil Mui, Huan Wang
Abstract
We present Salesforce Koa, an enterprise language model built by post-training the open-weight Nemotron-3-Super-120B foundation model with reinforcement learning using Group Relative Policy Optimization (GRPO). Salesforce Koa is trained on public and synthetically generated data, with no customer data, to improve tool use and agentic capabilities while preserving strong general-purpose performance. Its distinctive component is a simulation-to-reward pipeline that expands workflow specifications into persona-conditioned multi-turn tasks with task-resolution rewards grounded in successful tool use for data-dependent requests. For enterprise domains, these specifications are written in Agent Script, Salesforce's declarative language for building Agentforce agents; for public tool-use domains, we synthesize the workflow structure directly. The same simulation and grounded-reward machinery drives GRPO across both. Across public tool-use, agentic-reasoning, and enterprise Customer Relationship Management (CRM) benchmarks, Salesforce Koa improves over its open-weight base, with the clearest gains on multi-turn tool use, and surpasses a strong proprietary baseline while remaining below the strongest frontier models. These results show that specification-driven reinforcement learning is a practical path to specializing open-weight foundation models for enterprise agentic tasks.