Are Salesforce's AI Agents Its Most Important Innovation to Date?

The Gist

Agentforce and AI agents. Salesforce introduces autonomous AI agents that work alongside humans, potentially transforming knowledge work.
Tableau Einstein revolutionizes BI? Tableau unveils new capabilities to make data more accessible and actionable, aiming to increase data-driven decisions.
Marc Benioff's vision for AI. Salesforce's CEO shares insights on the future of AI, emphasizing the integration of AI agents in business operations.

SAN FRANCISCO — Salesforce and its Tableau subsidiary are setting the stage for a seismic shift in the future of knowledge work, especially in the areas of marketing, sales and support, with the introduction of "Agentforce” and Tableau Einstein.

Here's what's seismic about it: potential AI agents that don't just assist humans. They replace them. That's not a new concern for contact center agents, but it could be reality should AI technologies like the ones Salesforce revealed this week actually produce real results: the kind of results that don't make customers scream, "representative!" over and over.

Still, you'll find Salesforce's messaging all about a human and AI partnership in the world of customer service and support. Blasted across Salesforce's home page is the message, "Humans with Agents, driving success together."

So there's that. But beyond that marketing literature, where will the tech actually go?

These innovations — which will be unwrapped at the CRM provider's massive Dreamforce conference next week — promise to transform human roles, enhancing them with AI in some areas while potentially eliminating others, particularly in telesales and frontline support. At the same time as replacing some roles with AI, Agentforce aims to deliver more effective messaging that resonates with customers; so much so, that Salesforce CEO Marc Benioff said this is "the most important thing we've done at Salesforce."

Salesforce also unveiled for its Tableau product new capabilities that aim to revolutionize business intelligence, enabling users to access the data they need for faster, more informed decisions. At the heart of each of these changes is the convergence of generative AI and traditional model-based AI, driving a smarter, more agile workforce.

Agentforce Announcement: 'Third Wave of the AI Revolution'

Salesforce with Agentforce aims to usher in what Salesforce calls a "third wave of the AI revolution” with the introduction of AI agents that work alongside human employees to redefine how businesses operate.

Unlike earlier generations of AI, which were largely limited to providing insights or generating content, Salesforce says its agents act autonomously, taking over routine tasks and enabling human workers to focus on more complex, value-driven activities.

Salesforce argues that the evolution of AI can be seen in three distinct waves:

The first wave, predictive AI, used historical data to forecast future events.
The second wave, generative AI, powered by large language models (LLMs), created new content by understanding and mimicking human language, such as drafting marketing copy, generating images or crafting customer service emails.
The third wave, which Salesforce debuts this month, involves autonomous AI agents or "agentic systems," which not only suggest actions but reason through multifaceted projects without human oversight at every step.

This development aligns with the predictions of Klaus Schwab in "The Fourth Industrial Revolution," who foresaw the disruptive impact of AI and the subsequent need for governments to address the impending workforce displacement.

With this said, Salesforce claims its vision for this third wave of AI builds upon the concept of augmenting human capabilities, much as outlined by David De Cremer in "AI Savvy Leader."

However, Salesforce's approach in fact goes beyond merely augmenting human tasks — it seeks to create AI agents capable of eliminating entire categories of first-line functions. These customizable agents autonomously handle tasks that would otherwise require human intervention, freeing up the employees that remain to concentrate on responsibilities that require human judgment and creativity.

It is an open question to be fair whether these new capabilities will be resulting in less knowledge workers being needed. And how will future workers get trained with no entry level positions.

Related Article: Can Agentic AI Revolutionize CX and EX?

Clara Shih Introduces Low-Code Agent Builder for Custom AI Agents

Clara Shih, CEO of Salesforce AI, argues, “with the rise of autonomous AI agents — advanced digital colleagues that can take action without human intervention — that is starting to change.”

In her view, this represents a real paradigm shift where AI agents move from being tools that assist humans to entities that execute actions independently. Shih shared at the press briefing a low-code "agent builder" for customizing AI agents that can be integrated into every cloud service Salesforce offers. The agent builder allows users to define:

Role: Specifies what the agent does.
Data: The information the agent needs to perform its role.
Actions: The capabilities of the agent, including business logic, workflows and integrations. It enables existing business processes with AI agents.
Guardrails: Defines what the agent should not do, including security measures, data masking and human oversight.
Channel: Specifies where the agent performs its tasks, such as a specific platform or communication medium.

She is clear agents are designed to perform actions based on predefined workflows, or "flows," to enhance automation and efficiency across Salesforce's offerings. The cost savings are particularly valuable in healthcare. For example, an AI-driven agent handling a conversation for a company like Saks costs about $2 per call. This is notable because a human agent typically handles 10-20 calls per hour, making the AI option more cost-effective than the fully burdened cost of a level-one human agent.

AI Agents as Partners, Not Competitors

Salesforce interestingly suggests that we shouldn’t view AI agents as competitors to human workers, but as partners.

It's people with AI — not people versus AI, Salesforce officials will say. It's a way of bringing complementary skills and intelligence to the table to get even better business outcomes.

While today's AI agents may be akin to interns, Salesforce envisions a future where they operate more like full-time employees, with sophisticated capabilities and foundational knowledge.

Use Cases of Agentforce

Salesforce has already created over a hundred out-of-the-box autonomous AI use cases, including service reply to recommendations and sales account summaries. Some notable examples include:

Service Agent: Engages with customers 24/7, addressing inquiries, managing objections and scheduling meetings using CRM and external data, allowing human sellers to build deeper relationships.
Sales Coach: Provides personalized role-playing sessions for sales teams, leveraging Salesforce data and generative AI to tailor practice pitches and responses for specific deals.
Merchant: Assists ecommerce teams with site setup, goal setting, promotions, product descriptions and data insights to simplify daily operations.
Buyer: Enhances the B2B buying experience by helping buyers find products, make purchases and track orders via chat or sales portals.
Personal Shopper: Acts as a digital concierge, providing personalized product recommendations and search assistance.
Campaign Optimizer: Automates the full marketing campaign lifecycle by analyzing, generating, personalizing and optimizing campaigns based on business objectives.

Marc Benioff: CEO's Vision for the Future of AI

Envisions AI as a Transformative Collaborative Agent

Marc Benioff, CEO of Salesforce, envisions a future for artificial intelligence (AI) that is as transformative as it is unpredictable. "We are seeing things we never expected to see," he noted, underscoring the rapid evolution of AI and its ability to yield results even its creators can't fully explain.

Benioff highlights a striking example: a banking CEO unable to explain an unexpected AI-driven result. This points to a critical shift — from AI as a mysterious force to AI as a collaborative agent that works alongside humans to drive customer success.

Beyond Traditional AI Models: The Rise of Embedded Agents

At the heart of Benioff's vision is a move beyond traditional AI models like large language models (LLMs) and retrieval-augmented generation (RAG). He suggests that AI's real power lies not in these tools alone but in the rise of "agents" — AI systems embedded within businesses to transform how they connect with their customers.

Learning Opportunities

Webinar

Sep

CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms

Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.

Webinar

Sep

Call Spoofing Trends Financial Institutions Can't Afford to Ignore

Don't let robocalls sabotage your customer relationships. Learn how to secure your voice channel.

Webinar

On demand

Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential

Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.

Watch Now

Webinar

On demand

CX Reality Check: What Your Customers Are Really Thinking

Watch Now

Webinar

On demand

Content Leaders Collective: What a Good CCMS Actually Looks Like

Discover how content leaders transform clunky systems into growth engines — and why your competitors are ahead.

Watch Now

Webinar

On demand

From Hype to High-Impact CX Strategies That Actually Scale

Turn buzzworthy AI and outsourcing trends into measurable CX wins with fresh 2025 data.

Watch Now

Webinar

Sep

CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms

Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.

Webinar

Sep

Call Spoofing Trends Financial Institutions Can't Afford to Ignore

Don't let robocalls sabotage your customer relationships. Learn how to secure your voice channel.

Webinar

On demand

Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential

Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.

Watch Now

According to Benioff, 90% of AI's value will come from enhancing front-office operations. This approach is exemplified by Salesforce's Data Cloud, which offers better data access than competitors like Snowflake and Databricks, all without requiring customers to be experts in AI or install new systems, he promised.

From Salesforce to Agentforce: Making AI Accessible to All

Benioff believes the future of AI is a "platform approach," where businesses can simply turn on AI capabilities for their customers without installing new technology. This vision is already playing out in industries from healthcare to retail. At Kaiser Permanente, for example, AI agents help automate patient interactions, giving doctors more time for critical tasks. Disney uses AI to enhance customer experiences at its theme parks, driving more revenue by better understanding and responding to visitor needs.

Ultimately, Benioff argues that AI cannot be a "bolt-on" to existing systems — it must be integrated deeply within a company's operations. By working closely with customers to build out their own "agent forces," Salesforce aims to help businesses navigate this "third wave" of AI.

For Benioff, this isn't just a technological evolution; it's a moment of profound change, driven by an idealistic vision to make organizations more efficient, responsive and customer-centric. In his words, it's "the most important thing we've done at Salesforce." There are even whispers of people asking the CEO to rename the company Agentforce.

Introducing Tableau Einstein

According to Tableau's State of Data and Analytics Report, only 30% of enterprise employees currently use data to make business decisions. This low adoption rate reflects a failure to make data accessible and actionable for business users, a challenge that Southard Jones, chief product officer at Tableau, attributes to the proliferation of dashboards that did not measure up to their ambitious goals.

While dashboards have value, Jones believes they fall short in performance management and fail to drive business behavior. In many cases, they are supported by a collection of static, back-office reports. This is sad given the goals of balanced and linked scorecards.

The above deficiencies has motivated the launch of Tableau Einstein, signaling a shift in the analytics and BI marketplaces. The new focus is on meeting users where they already work—integrating insights directly into platforms like Slack or Salesforce.

Here, Tableau’s aims to enable marketers, sales and service professionals to combine data from inside and outside their organizations, utilizing AI and generative AI to deliver both pre-built and ad hoc analytics. The goal is to suggest measures and metrics to knowledge workers needing to know the why behind a piece of data. For these users, predictive insights are provided that with workflow transform data into meaningful actions, much like a Google Maps experience.

The Einstein approach is also designed to address issues of data trust and intuition. Tableau's new AI-driven connective layer facilitates rapid access to both structured and unstructured data in real-time, leveraging the Salesforce Data Cloud, which includes a vector database for sentiment analysis. This composable analytics strategy enables data packaging and sharing, making it easier for analysts to turn raw data into useful insights.

The integration of Tableau's capabilities with Salesforce further supports the "Agentforce" initiative, creating a workflow engine that turns insights into autonomous actions. The aim is to empower sales, marketing and service professionals by embedding real-time, context-driven analytics across platforms, driving smarter, faster decision-making at every level of the enterprise.

Salesforce's Future Vision for AI

Organizations like Bombardier, Heathrow Airport, and Wyndham are already leveraging Salesforce's AI capabilities, with companies like OpenTable leading the charge in deploying its "Agentforce" platform.

Importantly, Salesforce acknowledges that these powerful AI agents require strict guardrails. Salesforce provides tools for businesses to enforce data security, data privacy and policy compliance, ensuring that only authorized individuals can access specific data and workflows.

Salesforce’s vision for the future of AI agents is one of synergy and partnership between human and machine intelligence, promising a new era of productivity and innovation in the workplace.

Have a tip to share with our editorial team? Drop us a line: