Salesforce is today announcing the general availability of Agentforce 360, the latest iteration of its flagship AI product, as Dreamforce gets underway.
The CRM giant said that, while others are “just beginning” to showcase what’s possible with agentic AI, they have spent the past year delivering it, with four major releases, thousands of customer deployments, and Salesforce itself as Customer Zero.
Marc Benioff, Chair and CEO, Salesforce, said: “We’re entering the age of the Agentic Enterprise – where AI elevates human potential like never before. Agentforce 360 connects humans, agents, and data on one trusted platform, helping every employee and every company achieve more than they ever thought possible.”
Let’s take a look at what exactly makes Agentforce 360 different from its predecessors.
Agentforce 360 Features Explained
During Agentforce’s year-long history, it has gone through several iterations, including:
- Agentforce (October 2024): The beginning of Salesforce’s enterprise AI agent platform.
- Agentforce 2 (December 2024): Improving the Atlas Reasoning Engine for more predictable, grounded results.
- Agentforce 2dx (March 2025): Allowing agents to be embedded in any workflow – proactive, triggered, and cross-functional.
- Agentforce 3 (June 2025): Providing enhanced interoperability and governance, preparing enterprises for scale.
The latest version, which Salesforce are calling ‘Agentforce 360’, is a combination of four ‘ingredients’:
- Agentforce 360 Platform: The foundation of Agentforce, now featuring a new conversational builder, hybrid reasoning for greater control and accuracy, and voice capabilities.
- Data 360 (aka Data Cloud): The data layer that gives agents context. With innovations like Intelligent Context and Tableau Semantics, companies can turn unstructured data and analytics into rich context and understanding for AI.
- Customer 360 Apps: The business logic and institutional memory of every enterprise – capturing how it sells, serves, markets, and operates – now brought to life through AI agents that deeply understand every customer and process from the inside out.
- Slack: The conversational interface for humans and agents to work together, connecting knowledge, actions, and data in real time.
This, Salesforce says, enables businesses to deploy agents grounded in governed, trusted data; work across teams and workflows; collaborate with humans and other agents directly in Slack; and leverage existing processes, business logic, and data infrastructure.
And with its open ecosystem, partners extend Salesforce’s technology into “every industry with tailored solutions and services”, the cloud giant says.
So, what’s new in Agentforce 360?
Agentforce Platform
Agentforce 360 aims to introduce the most dependable and accessible agents yet – designed to reason, act, and collaborate alongside people, with features including:
- Agentforce Builder: A new conversational development studio that lets teams design, test, and deploy agents using natural language, with no manual configuration.
- Agentforce Voice: A native voice layer that transforms IVR systems into natural, real-time conversations with low-latency transcription, realistic speech synthesis, and deep Salesforce integration.
- Hybrid Reasoning and Agent Script: This combines deterministic workflows with flexible LLM reasoning. Developers can define guardrails, tool use, and logic using Agent Script, powered by the configurable Atlas Reasoning Engine.
- Agentforce Vibes: Extends low-code development to AI, letting builders “vibe-code” apps grounded in company data and governance.
- Observability: New dashboards help teams monitor reasoning, accuracy, and compliance.
The Agentforce 360 Platform is built on learnings from more than 12,000 implementations.
Agent Script – a new scripting language for controlling agents – is a core component of the new release, allowing you to define agent behavior with a human-readable expression language, allowing for conditional logic, precise tool use, and guided, deterministic controls.
Agent Script is a high-level, human-readable, portable JSON expression language designed for controlling agents, allowing teams to define complex agent logic, including conditionals, “if/then” rules, and hand-offs.
The Atlas Reasoning Engine is also now configurable, allowing teams to “balance the inherent creativity” of LLMs with the “certainty of structured business logic”, Salesforce says, ensuring more reliable agent behavior.
Agentforce 360 Platform also now supports Google Gemini as a model for the Atlas Reasoning Engine, joining OpenAI and Anthropic to allow even more choice for what models Agentforce uses to reason and plan.
Meanwhile, Agentforce Voice delivers natural, human-like conversations that “capture your brand’s unique tone and personality”.
Salesforce said in a statement: “We put each agent through thousands of real-world scenarios – including noisy backgrounds, distracted callers, and complex, multi-step processes – to ensure reliable, enterprise-ready performance. The result? Stronger customer relationships and measurable ROI.”
Key capabilities include:
- Fully personalized voice agents: Each agent speaks in a voice aligned to the company’s brand, trained on its data, tone, and contextual knowledge.
- Ultra-realistic, low-latency interactions: Built for real-time responsiveness, Salesforce says that Agentforce Voice achieves “near-human conversational flow” while executing actions like updating CRM records, triggering workflows, or calling APIs – with full auditability through Data 360.
- Seamless integration: Natively connected to Salesforce, Agentforce Voice extends across the contact center stack with support for Amazon Connect, Five9, NiCE, and Vonage.
- Fluid, interruptible conversations: Every word is transcribed live in Salesforce Voice, giving humans real-time visibility from within the agent console to allow instant takeovers when needed.
- Voice-first platform architecture: Enhanced latency, speech realism, and deep language understanding.
The Agentforce Builder means teams can craft, test, and refine agents in a single, conversational workspace.
Salesforce says that, whether you’re a seasoned developer or a natural language builder, you can ditch the separate build, test, and deploy cycles, and revolutionize your workflow with “vibe-building,” instant editing in doc-like, canvas, or script views, and one-click simulations with real-time debugging.
Every agent compiles into a portable JSON file, making versioning and sharing simple, while enterprise-grade security and privacy controls ensure top-tier governance.
Key capabilities include:
- Unified workspace: Collapse drafting, testing, and deployment into a single interface, eliminating the traditional build-test loop.
- Multi-view editing: Switch seamlessly between a doc-like editor with autocomplete, low-code canvas, and a script view to match any workflow.
- Rapid iteration with simulation and live debugging: Preview real interactions in real time, view trace data for each reasoning step, and adjust instantly.
- Enterprise-grade governance: Built on Salesforce’s security and privacy features to ensure compliance and safe deployment.
- Agentforce Vibes: An AI coding partner that collaborates like a pair programmer, understanding your project context and executing on your behalf to accelerate development.
With Intelligent Context, AI agents are grounded in the complex, unstructured reality of a business, not just in structured data, Salesforce says. It pulls critical information from everything from PDFs and tables to images and flowcharts.
Powered by the new ‘Data 360’ – more on this below – this feature automatically extracts, structures, and surfaces this information so agents can deliver accurate, relevant, and business-specific answers.
Salesforce claims their low-code pipeline lets you configure document processing in hours instead of days, and you can configure context so agents can interpret the same content from multiple business perspectives.
Key capabilities include:
- Automated context extraction: Converts complex, unstructured documents into structured, actionable data for Agentforce.
- Low-code configuration: Rapidly set up unstructured data pipelines in hours, not days, without extensive engineering effort.
- Multimodal support: Handles PDFs, tables, images, flowcharts, and other diverse content types.
- Seamless integration with Data 360: Create a search index with AI, then activate it everywhere for secure, contextual, enterprise-ready operations.
Data 360 (Formerly Data Cloud)
In the press material outlining the new features, the phrase ‘Data 360’ is hyperlinked to salesforce.com/data – the homepage for Data Cloud. So, it seems like another Salesforce product renaming!
With Agentforce 360, Data 360 activates a company’s structured or unstructured data to provide every agent with business context and personalization, Salesforce says.
Intelligent Context enables agents to access unstructured content, such as PDFs or diagrams, to guide users through real-world tasks, including troubleshooting and analysis.
Tableau Semantics translates data into business language and ensures consistent metrics across clouds through the Customer 360 Semantic Data Model (SDM).
“With partners Databricks, dbt Labs, and Snowflake, Salesforce is enabling standardized semantics across platforms,” the company says.
Customer 360 Apps
Salesforce apps now embed conversational agents that can act directly in business workflows.
- Agentforce Sales: This automates prospecting, qualification, and quoting with next-best actions.
- Agentforce Marketing: This builds and launches campaigns, journeys, and assets autonomously.
- Agentforce Service: Powering the Command Center for Service for proactive, always-on support.
- Agentforce Field Service: Automates scheduling, integrating live maps via Esri, and enabling hands-free data capture with Voice to Form.
- Agentforce Revenue Management: Drives predictable growth with AI agents that empower sales, ops, and billing teams with agentic quoting, agentic billing, and consumption management.
- Agentforce Commerce: Helps increase shopper conversion and grow cart size with conversational, guided shopping experiences and checkout.
- Agentforce IT Service: Helps reduce IT costs and boost employee productivity by ending the “portal-to-ticket era” and deploying specialized AI agents to deliver 24/7 conversational resolutions everywhere employees work.
- Agentforce 360 for Industries: Delivers faster time to value with new pre-built, industry-specific solutions like Agentforce Life Sciences, Agentforce Public Sector, and Agentforce Manufacturing.
Slack: The ‘Agentic OS for the Enterprise’
Salesforce says Slack has become the “system where humans, agents, apps, and data connect in real time”.
- Slack-First Apps: Agentforce Sales, IT Service, HR Service, and Tableau Next now surface insights and complete actions without leaving Slack.
- Channel Expert Agent: Always-on expertise inside channels, powered by real-time enterprise knowledge search.
- Enterprise Search: Natural language answers spanning Google Drive, GitHub, Jira, and more.
- Reimagined Slackbot: Context-aware assistant for writing help, message summaries, and huddle notes.
- Model Context Protocol (MCP): Integrates third-party AI like Anthropic, Dropbox, and OpenAI through new APIs for real-time knowledge and automation.
According to The Verge, Slack is testing an update for Slackbot – which currently works as a reminder and notification tool – to turn it into an AI assistant, with the ability to create custom plans tailored to a workspace, sift through messages, and gather information.
Rob Seaman, Chief Product Officer of Slack, told The Verge: “Slackbot today is fairly rudimentary. But what we’ve done is we’ve actually rebuilt it from the ground up as a personalized AI companion.”
Slackbot will appear as an icon next to the search bar as part of the pilot, and when it’s clicked, a DM-style panel will open up where the user can input prompts like ‘What are my priorities for today?’
Agentforce 360 is available globally today, with continued innovation arriving in pilot and beta over the coming months.
Final Thoughts
Another Dreamforce, another Agentforce!
Last year’s conference was very much Agentforce-themed, serving as the official launch of the product, and the ecosystem will likely not be surprised that this year’s big reveal is Agentforce-themed.
The post Salesforce Unveils ‘Agentforce 360’ at Dreamforce ’25 appeared first on Salesforce Ben.