Salesforce bets on AI 'agents' to fix what it calls a $7 billion problem in enterprise software
While 50,000 visitors use Salesforce Dreamforce Conference this week, the enterprise software giant is making its most aggressive bet yet on artificial intelligence agents, positioning itself as the antidote to what it calls an industry-wide “pilot purgatory” where 95% of business AI projects never reach production.
The company launched on Monday Agent Force 360a major reinvention of the entire product portfolio, designed to transform companies into what they call “agentic enterprises”: organizations where AI agents work alongside humans to handle up to 40% of sales, service, marketing and operations work.
“We’re really in the age of agentic AI, and I think this is probably the biggest revolution, the biggest technology transition that I’ve ever seen in my career,” Parker Harris, co-founder and Chief Technology Officer of Salesforce, said at a recent press conference. “In the future, probably 40% of the work in the Fortune 1000 will be done by AI, and humans and AI will actually work together.”
The announcement comes at a pivotal time for Salesforce, which has deployed more than 12,000 AI agent deployments in the past year while building what Harris called a “$7 billion company” around its AI platform. Yet the launch also comes amid unusual turbulence, as CEO Marc Benioff experiences fierce reaction for recent comments support for President Trump and suggests that National Guard troops should patrol the streets of San Francisco.
Why 95% of business AI projects never get started
The stakes are enormous. While companies rushed to experiment with AI following the rise of ChatGPT two years ago, most enterprise implementations have stalled before reaching production, according to recent MIT research which Salesforce executives have cited extensively.
“Customers have invested a lot in AI, but they’re not getting the value from it,” said Srini Tallapragada, president and chief engineering and customer success officer of Salesforce. “95% of enterprise AI pilots fail before production. It’s not for a lack of intent. People want to do this. Everyone understands the power of the technology. But why is it so hard?”
The answer, according to Tallapragada, is that AI tools are still disconnected from business workflows, data and governance systems. “You write prompts, prompts, you get frustrated because the context isn’t there,” he said, describing what he called a “prompt doom loop.”
Salesforce’s solution is a deeply integrated platform that connects four ingredients: the Agentforce 360 agent platform, Data 360 for unified data access, Customer 360 apps with business logic, and Slack as the “conversational interface” where people and agents collaborate.
Slack becomes the front door of Salesforce
Perhaps the most important strategic shift is the increase of Slim — acquired by Salesforce in 2019 for $27.7 billion – as the primary interface for Salesforce itself. The company is reimagining its traditional Lightning interface around Slack channels, surfacing sales deals, service cases, and data insights through conversations instead of forms and dashboards.
“Imagine maybe you don’t log into Salesforce, you don’t see Salesforce, but it’s there. It comes to you in Slack because that’s where you get your work done,” Harris explains.
The strategy includes embedding Salesforce’s Agentforce agents for sales, IT service, HR service, and analytics directly into Slack, in addition to a completely rebuilt Slackbot that acts as a personal AI companion. The company is also launching “Channel expert,” an agent who is always available and answers channel calls immediately.
To give third-party AI tools access to Slack’s conversation data, Salesforce is releasing a Real-time search API And Model Context Protocol Server. Partners such as OpenAI, Anthropic, Google, Perplexity, Writer, Dropbox, Notion, and Cursor are building agents that will live natively in Slack.
“The best way to see the power of the platform is through the AI apps and agents that are already being built,” Rob Seaman, a Salesforce director, said during a technical briefing, citing examples of startups “reaching tens of thousands of customers who installed it in 120 days or less.”
Voice and IT services are targeting new markets
In addition to the Slack integration, Salesforce has announced major expansions in voice-based interactions and employee services. Agentforce voicenow widely available, transforms traditional IVR systems into natural conversations that can update CRM records, trigger workflows and seamlessly hand off to human agents.
IT service offerings represent Salesforce’s most immediate challenge ServiceNowthe market leader. Mudhu Sudhakar, who joined Salesforce two months ago as senior vice president of IT and HR Service, positioned the product as a fundamentally new take on employee support.
“Legacy IT service management mainly consists of portals, forms, tickets and manual processes,” says Sudhakar. “We had a few key principles: conversation first, agent first, really focused on having a conversational experience for the people asking for support and for the people providing the support.”
The IT services platform includes what Salesforce describes as more than 25 specialized agents and more than 100 pre-built workflows and connectors that can handle everything from password resets to complex incident management.
Early customers report dramatic efficiency gains
Customer results show that the approach is gaining more and more ground. Reddit Average support resolution time was reduced from 8.9 minutes to 1.4 minutes – an 84% improvement – while 46% of cases were referred entirely to AI agents. “These efficiencies have allowed us to provide on-demand assistance for complex tasks and increase advertiser satisfaction scores by 20%,” John Thompson, Reddit’s VP of Sales Strategy and Operations, said in a statement.
Enginea travel management company, has reduced average handling time by 15%, saving over $2 million annually. OpenTable resolved 70% of questions about restaurants and dining autonomously. And 1-800Accountant achieved a 90% deflection rate during the critical tax week period.
Salesforce’s own internal implementations are perhaps the most telling. Tallapragada’s customer success organization now handles 1.8 million AI-powered calls every week, with statistics published at help.salesforce.com show how many officers respond versus escalate to people.
More importantly, Salesforce has deployed AI-powered sales development representatives to follow up on leads that previously would have gone uncontacted due to cost constraints. “Now Agentforce has an SDR that follows up on thousands of leads,” Tallapragada explains. The company also increased proactive customer outreach by 40% by shifting staff from reactive support.
The trust layer problem that companies cannot ignore
Given enterprise concerns about the reliability of AI, Salesforce has invested heavily in what it calls the “trust layer”: audit trails, compliance checks, and observability tools that allow organizations to monitor agent behavior at scale.
“You have to think of an agent as a human. Digital labor. You have to manage performance just like a human. And you need these audit trails,” Tallapragada explains.
The company faced this challenge when scaling up its use of in-house agents. “When we started at Agentforce at Salesforce, we tracked every message, which is great up to 1,000, 3,000,” says Tallapragada. “Once you have a million chats, there’s no human left, we can’t do it anymore.”
The platform now includes “Agentforce grid” for searching millions of conversations to identify and resolve problematic patterns. The company also introduced Agent Script, a new scripting language that allows developers to define precise guardrails and deterministic controls for agent behavior.
The data infrastructure is getting a major upgrade
Underpinning agent capabilities is significant infrastructure investment. The one from Salesforce Data 360 includes ‘Intelligent Context’, which automatically extracts structured information from unstructured content such as PDFs, diagrams and flowcharts using what the company describes as ‘AI-powered unstructured data pipelines’.
The company also collaborates with Databricks, dbt LabsAnd Snowflake about the ‘Universal Semantic Interchange’, an attempt to standardize how different platforms define business metrics. The upcoming $8 billion acquisition of Informatica, expected to close soon, will expand metadata management capabilities across the enterprise.
The competitive landscape is becoming increasingly intense
Salesforce’s aggressive AI agent push comes as virtually every major enterprise software vendor is pursuing similar strategies. Microsoft has integrated Copilot into its product line, Google offers agent capabilities through Vertex AI and Gemini, and ServiceNow has launched its own agent offering.
When asked how Salesforce’s announcement compares The recent releases of OpenAITallapragada emphasized that customers will use multiple AI tools simultaneously. “Typically I see they use OpenAI, they use Gemini, they use Anthropic, like Salesforce we use all three,” he said.
The real difference, executives reasoned, lies not in the AI models, but in the integration with business processes and data. Harris described the competition in terms that were familiar from Salesforce’s founding: “26 years ago we just said, let’s make Salesforce automation as easy as buying a book on Amazon.com. And we are. We want to make agentic AI as easy as buying a book on Amazon.”
The company’s customer success stories are impressive, but still represent a small portion of its customer base. With 150,000 Salesforce customers and a million Slack customers, the 12,000 Agentforce deployments represent a penetration rate of about 8% – strong for a one-year-old product line, but hardly ubiquitous.
The company’s shares have fallen so far about 28% immediately Relative strength of only 15suggests that investors remain skeptical. This week’s Dreamforce demonstrations – and the months of customer deployments to follow – will begin to provide answers to the question of whether Salesforce can finally move AI from pilots to large-scale production, or whether the “$7 billion business” remains more ambition than reality.
