Our VML team has just returned from Dreamforce 2025. Dreamforce is the biggest vendor tech events with 40,000 to 50,000 in-person attendees and over 200,000 people registered to attend online. Across just three days, more than 1.500 sessions took place! 

To help you navigate this wealth of content, we’ve curated a list of recommended sessions and highlights tailored to each area of expertise below. Feel free to explore and dive into the insights that matter most to you.

If you have any questions regarding the material, feel free to reach out to to [email protected] or [email protected]

Jump straight to your favorite content

Salesforce revealed its vision for the future moving away from the traditional “Clouds” naming. And yes, it’s packed with Data and AI.

To stay competitive, Salesforce is transforming every major function - marketing, service, commerce - into what they’re calling an Agentic Enterprise (buzzword alert!).

The product names have changed once again: no more Marketing Cloud or Sales Cloud. From now on, it’s Agentforce Marketing, Agentforce Sales, and so on. Meanwhile, the CDP Data Cloud has been rebranded as Data 360 - the foundation for all things agent context.

There’s a strong emphasis on Data 360 (formerly Data Cloud) as the backbone for building AI agents that deliver control, predictability, and consistent performance - especially for mission-critical work.

Dreamforce Main Keynote: Become an Agentic Enterprise

Understand the four major components necessary to build agents that perform with control and consistency, focusing on context, experiences, and observability.

Watch the complete version (~1.5 h) 

Watch the short version (~6 min)

Salesforce once again emphasized that the new and old Marketing Clouds will continue to coexist and stay connected. For instance, you’ll still be able to trigger “old” journeys from within the new Marketing Cloud. The keynote also showcased how traditional one-way marketing can evolve into adaptive, two-way conversations.

There were also many sessions on Data 360 (Salesforce’s CDP), focused on creating a unified customer profile from disparate systems. We saw a clear roadmap for Marketing Cloud users to tap into this unified data for personalization - without needing a massive migration.

The product isn’t fully there yet, but give it 6–12 months, and it’s likely to mature significantly

Marketing That Sparks a Conversation | Marketing Keynote, Dreamforce 2025

Salesforce wanted to show the audience how easy it is to connect existing Marketing Cloud data and map complex data sets into Data 360 to immediately activate two-way messaging and better personalization. In reality, we know that it requires a lot more analysis and alignment but Data 360 makes it easier than any other CDP.

Watch the keynote (~1h) 
 

Your Marketing Maturity Model: What AI to Use When | Dreamforce 2025

Understand the AI maturity path, focusing on low-complexity, high-impact tools like out-of-the-box models (e.g., send time optimization) to drive engagement.

Watch the session (~30 min) 


Activate Trusted Data Everywhere | Data Cloud Keynote, Dreamforce 2025

The Salesforce's Data Cloud Keynote recasts its rebranded "Data 360" platform as the essential solution to the data fragmentation hindering enterprise AI. With new tools like "Data 360 Clean Rooms" and "Intelligent Context", the core message is clear: trustworthy AI depends entirely on first solving this underlying data mess.

Watch the session (~30 min)

We saw how AI agents act as digital store associates, handling everything from personalized product recommendations to purchase inquiries. This push toward Agentic Unified Commerce connects online and in-store operations to provide seamless, end-to-end customer experiences that strive to give websites the "same warm feeling" as physical stores.

Drive Growth with Agentforce Commerce | Commerce Keynote, Dreamforce 2025

Watch the full demo showing how a Commerce agent facilitates a complex purchase, including contextual product recommendations and purchase completion, turning conversation into conversion.

Watch the session (~1h) 

Unlock Growth with Agentic Brands | Consumer Goods Keynote, Dreamforce 2025

See how AI agents are transforming retail execution by proactively identifying issues (e.g., cooler temperature being too high) and automatically creating service cases to ensure optimal store conditions.

Watch the session (~45 min)


Cut B2B Commerce Costs with AI and Extensibility | Dreamforce 2025

See how developers used just four natural language prompts with AI to execute complex component customizations and deployment, estimating a saving of approximately $10,000 and four weeks of development time.

Watch the session (~20 min) 

We noted a strong focus on building an open, AI-powered, and developer-friendly architecture for Marketing, Commerce and Service.

We observed that Salesforce is working on eliminating the "decision gap" between insight and action. Agentic Analytics uses a unified AI Semantic Layer (in Tableau Next and Data 360) to transform complex structured and unstructured data (like call logs and emails) into contextual, actionable insights.

Powering Decisions, Driving Actions | Tableau Keynote, Dreamforce 2025

Watch the demonstration showing how investigating a stalled opportunity in a dashboard leads to immediate action by triggering a Salesforce Flow and a Slack notification, closing the decision gap.

Watch the keynote (~50 min) 


Embed Analytics and Drive Actionability with Tableau Next | Dreamforce 2025

Accelerate Semantic Modeling with Data Pro and Built-in AI

Explore AutoSDM (Auto Semantic Model creation), which uses natural language description to automatically generate relationships and calculated fields for semantic modeling, saving days of manual effort.

Watch the session (~20 min) 


Zero Copy 101: Building Open Data Ecosystems with Data Cloud | Dreamforce 2025

Understand the Zero Copy blueprint and how it allows Data 360 to access and analyze data directly from external hyperscalers (like Snowflake or Databricks) without duplicating the data.

Watch the session (~20 min)

This goes beyond ChatGPT and n8n. Why? Because scaling AI requires robust architecture, not just prompts and a loop. This track focused on the how of production AI, highlighting Agentforce 360 and tools like the new Agent Builder and Agent Script for deterministic control over complex business logic. We saw that ensuring agent predictability and incorporating observability tools are essential for managing a successful digital workforce, moving past the 95% failure rate of AI pilots.


Use Words, Not Code with Agent Builder | Dreamforce 2025

Dive into the new Agent Builder and Agent Script, which allows conversational agent creation while embedding deterministic, predictable control using conditional logic (if/then statements).

Watch the session (~15 min)

Deep Dive into Agent Performance and Analytics | Dreamforce 2025

Explore the foundation of Agentforce Observability, including the session tracing data model and the concept of breaking down interactions into "moments" for accurate quality scoring and optimization.

Watch the session (~15 min)

Do you have any questions?

Jaap van Oort

Jaap van Oort

Strategy Director CXM

Contact Jaap
Dienand Christe

Dienand Christe

Executive Director 

Contact Dienand

Please provide your contact information to continue.

Before submitting your information, please read our Privacy Policy as it contains detailed information on the processing of your personal data and how we use it.

Related Content

IMG 8150
Insight

You’re not as customer-centric as you say you are.

The steps towards a better customer experience
Read Article
VML Logo

Unified Digital Experiences

VML Session: Transforming Campaign Operations
Read More