Salesforce unveils Agentforce 3 to accelerate enterprise AI adoption
Salesforce has released Agentforce 3, an upgrade to its enterprise AI agent platform designed to address key challenges faced by organisations scaling artificial intelligence implementation.
The latest version introduces features that enhance visibility and control for enterprise users, moving AI agents from experimental use to full-scale deployment. More than 100 prebuilt actions, expanded interoperability options, and a centralised suite of tools for managing hybrid workforces are among the major enhancements.
Agentforce Command Centre
Agentforce 3 introduces the Command Centre, marketed as the first real-time observability suite specifically for hybrid workforces. It provides business leaders with a unified interface to monitor performance metrics across both AI agents and human team members. This monitoring is integrated into business key performance indicators (KPIs), aimed at supporting data-driven management and optimisation of workflow processes.
The Command Centre includes dashboards detailing agent adoption, feedback, success rates, costs, and specific topic performance, enabling teams to analyse usage trends and identify areas for improvement. The system records all agent activity in a session-tracing data model leveraging Data Cloud and is compatible with monitoring tools such as Datadog, Splunk, and Wayfound through the OpenTelemetry standard.
Teams and supervisors can also receive live analytics on latency, escalation frequency, and error rates with real-time alerts for anomalies, enabling prompt interventions to maintain service levels. For contact centres, agent activity can be monitored in real time, permitting quick performance escalations.
Interoperability and Prebuilt Integrations
Agentforce 3 expands interoperability with native support for open standards like Model Context Protocol (MCP) and Agent-to-Agent Protocol (A2A). The platform also integrates with more than 30 partners, including services such as Stripe, Box, Atlassian, AWS, Cisco, Google Cloud, IBM, Notion, PayPal, Teradata, and WRITER via its expanded AgentExchange.
Built-in interoperability is intended to allow AI agents not only to communicate with one another, but with broader enterprise applications, creating a more connected digital workplace.
Deployment and Usability
The update aims to reduce deployment times, with over 100 new prebuilt industry actions and flexible pricing designed to enable companies to deploy agents within days rather than months. The new architecture underlying Agentforce 3 promises improvements in accuracy, latency, resiliency, and model control, the latter of which addresses the requirements of regulated industries.
Agentforce Studio is a central space for building, testing, and optimising agents with the assistance of real-time analytics. It provides tools to simulate agent behaviour at scale and generate topics or test cases using natural language, with the goal of assisting teams in developing and deploying agents more efficiently.
Customer Adoption
Several organisations have already adopted Agentforce, including New Zealand-based companies such as Fisher & Paykel, Urban Rest, and Farm Focus, which cite benefits including task automation, enhanced customer engagement, and the ability for staff to focus on complex, people-centred work.
According to Salesforce, Engine, Grupo Globo, PepsiCo, UChicago Medicine, and 1-800Accountant have reported measurable improvements: Engine reduced its average customer case handling time by 15%, 1-800Accountant autonomously resolved 70% of administrative chat engagements during the peak tax season, and Grupo Globo saw a 22% increase in subscriber retention.
Adam Evans, Executive Vice President and General Manager of Salesforce AI, commented on the platform's development trajectory:
"With Agentforce, we've unified agents, data, apps, and metadata to create a digital labour platform, helping thousands of companies realise the promise of agentic AI today," said Adam Evans, EVP & GM of Salesforce AI.
Evans added:
"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."
Ryan Teeples, Chief Technology Officer at 1-800Accountant, described their experience with the system:
"Agentforce autonomously resolved 70% of 1-800Accountant's administrative chat engagements during the peak this past tax season, an incredible lift during one of our busiest periods. But that early success was just the beginning. We've established a strong deployment foundation and weekly are focused on launching new agentic experiences and AI automations through Agentforce's newest capabilities. With a high level of observability, we can see what's working, optimize in real time, and scale support with confidence," said Ryan Teeples, Chief Technology Officer at 1-800Accountant.
AI Agent Adoption Trends
Salesforce cited a soon-to-be-released Slack Workflow Index showing a 233% increase in AI agent usage in six months, with 8,000 customers signing up for Agentforce deployment during the same period. The platform is presented as a response to enterprise demands for improved governance, tooling, and observability in large-scale AI agent operations.
Agentforce 3 is now available globally, with tools and integrations focused on supporting enterprise clients as they expand AI adoption within their organisations.
