You Can Now Monitor and Scale AI Agents with Agentforce 3
Salesforce has launched its Agentforce 3, which is the biggest update to its AI agent platform that aims to give companies better oversight, faster development, and easier scaling of their digital agents.
The launch comes as the use of AI agents in enterprises continues to rise. According to a soon-to-be-published Slack report, usage has surged 233% in the last six months. But despite this growth, many businesses have struggled with visibility, governance, and control. Agentforce 3 is Salesforce’s answer to that gap.
The main highlight of this update is the new Command Center, which gives teams a central view of how their agents are performing. It tracks agent health, success rates, and costs and even flags problems in real-time so teams can fix them quickly. It also integrates with tools like Datadog, Splunk, and Wayfound, making it easier for teams to work in their existing monitoring setups.
With Agentforce, we’ve unified agents, data, apps, and metadata to create a digital labor platform, helping thousands of companies realize the promise of agentic AI today. Over the past several months, we’ve listened deeply to our customers and continued our rapid pace of technology innovation. The result is Agentforce 3, a major leap forward for our platform that brings greater intelligence, higher performance, and more trust and accountability to every Agentforce deployment. Agentforce 3 will redefine how humans and AI agents work together — driving breakthrough levels of productivity, efficiency, and business transformation.
Adam Evans, EVP & GM of Salesforce AI
Another major change is the support for Model Context Protocol (MCP), which allows companies to plug Agentforce into their current systems more easily. With native MCP support and tools from partners like MuleSoft and Heroku, businesses can turn APIs into agent-ready actions and manage them with built-in security and tracing.
Apart from that, the company is also expanding its AgentExchange ecosystem, giving customers access to AI-ready tools and services from over 30 partners, including AWS, Google Cloud, Stripe, PayPal, and WRITER. For example, companies can now let agents summarize documents from AWS, interact with Google Maps and Imagen models, or process payments via PayPal and Stripe.
Under the hood, Agentforce 3 is built on Salesforce’s updated Atlas architecture, which offers faster performance (with 50% lower latency), response streaming, inline citations for answers, and global availability in Canada, the U.K., India, Japan, and Brazil. Besides, the language support has expanded to six new languages, with more on the way.
When it comes to the pricing part, Agentforce 3 comes with flexible pricing, including unlimited usage SKUs for employee-facing agents in sales, service, and industry-specific workflows. There are over 100 new prebuilt industry actions included in this release, from patient scheduling to ad proposal generation. Agentforce is already available for government customers as well, with FedRAMP High authorization through Government Cloud Plus.
Want the latest tech news? Join our WhatsApp channel for quick updates!