2024 Sessions

Explore
Dreamforce 2024
sessions.

View 1,500+ sessions by role, topic, or industry to see an example of what's to come at Dreamforce 2025.

You are currently viewing sessions from Dreamforce 2024.
Back to Session CatalogBack to Session Catalog
Improve Agent Creation with Advanced Planner
Discover how to create next-generation agents by using Topics, Actions, and Advanced Planner. Enhance error handling, ensure accurate responses, and achieve seamless conversational flow.
Important: This session is first come, first seated. You do not have a reserved seat.

, Vice President Product Design and UX, AI Platform, Salesforce

, Conversation Design Senior, Salesforce

Role: Administrator, Designer
Industry: Technology
Product: Einstein, Data 360
Topic: Artificial Intelligence
Session Level: Intermediate
Session Type: Theater
  • Tuesday, Sep 171:00 PM - 1:20 PM PDT
    (8:00 PM - 8:20 PM GMT)
    Moscone West, L1 Forest, Redwood Theater
Session Summary by Einstein: This session focused on the fundamental building blocks of Agents, exploring recent upgrades Salesforce made to make Agents more accurate, efficient, and capable of delivering better outcomes. The new Agentforce platform introduces improved planning capabilities through fundamental building blocks like topics, descriptions, instructions, and actions that streamline agent responses. Topics create boundaries for agents, allowing them to focus on the user's intent, while descriptions guide the Large Language Models (LLMs) in topic selection, ensuring relevance and scope. Instructions and actions then dictate the specific tasks and steps the LLM must execute to provide tailored, context-aware responses. The talk emphasized the importance of designing Agents with principles like inclusivity, transparency, and relevance in mind. Techniques for crafting effective agent prompts include using "must" and "strictly" to convey strong obligations for more authoritative instructions.
Key Takeaways by Einstein:

• Agentforce's new components like topics and descriptions help AI agents to maintain focus and context, improving the efficiency and capability of agents on the platform.

• Design principles such as inclusivity, transparency, and relevance are vital in creating agent experiences that cater to diverse users and build trust through clear and relevant interactions.

• Specific techniques for prompt engineering, like using "must" over "should" were shared to help attendees craft more effective instructions for their AI agents, ensuring precise agent behavior and desired outcomes.

Einstein Feedback : Report a Concern

With sessions and hands-on activities led by product experts, visionaries, and fellow Trailblazers, Dreamforce offers a deep dive into the latest in CRM + AI + Data + Trust.

Merlinda Lu

Salesforce Developer, IBM