AI is only as good as the value it delivers. See the framework Salesforce uses to track agentic impact, optimize response accuracy, and turn raw agent logs into actionable Tableau dashboards.
Sessions
Hundreds of sessions now live. More on the way.
This year’s lineup is stacked with hands-on trainings, product deep dives, and real-world strategies you can apply right away.
Register now and we’ll notify you when it’s time to build your agenda and reserve your seat.
New sessions are added daily and subject to change. Learn more about
session types. Hands-on training sessions coming soon!
Sessions
Snag your seat.
New this year, you can now reserve your seat at most sessions. Register now to get first-access to hands-on learning and expert-led sessions.
Tip: Use " " around your search terms to find an exact phrase.
Tip: Use " " around your search terms to find an exact phrase.
Community Speaker
Reserved Seating
Video On Demand
Beyond the Hype: How We Measure AI at Salesforce
, Senior Director, Data and Analytics, Salesforce
, Principal Visualization Engineer, Salesforce
Role: Data Analyst, Product Management
Product: Tableau, Agentforce
Session Level: Beginner
Session Type: Theater
- Thursday, May 710:00 AM - 10:20 AM PDT(5:00 PM - 5:20 PM GMT)Theater 1 | Data Village, Ground LevelFIRST COME FIRST SEATED
Session Summary: This session provided a comprehensive overview of how Salesforce measures AI performance at scale using a robust, data-driven framework. With over 300 internal agents spanning 20 organizational units, the team developed a measurement system focused on four core pillars: engagement, efficacy, performance, and satisfaction. By leveraging Tableau dashboards, AI-curated narratives, and advanced metrics like re-prompt rate and negative response rate, Salesforce has been able to evaluate agent quality, track business value through time savings, and provide actionable insights to users. Additionally, the session highlighted the use of AI and machine learning models to scale evaluations and emphasized intentional design choices to create cohesive and trusted agent experiences across the organization.
Key Takeaways:
• Salesforce's measurement system evaluates AI agents on engagement, efficacy, performance, and satisfaction, using tools like Tableau and AI-augmented evaluation techniques.
• The business value of AI agents is quantified by analyzing the complexity and usage of agent actions, allowing Salesforce to determine time savings and operational impact across its portfolio of over 300 agents.
• Future developments will focus on understanding the interdependencies between agent metrics, referred to as 'chemistry,' and providing diagnostic tools for immediate, actionable insights to improve agent performance.
Session Summary Feedback : Report a Concern
Available Now on Salesforce+: Watch now
