Agentforce Basics
Understanding the Basics of Agentforce:
Agentforce revolutionizes processing efficiency with its Test Center, designed to address the limitations of Agent Builder, which requires testing one interaction at a time. By leveraging batch testing via CSV uploads, you can now:
- Upload up to 10 batches within 10 hours.
- Test up to 200 cases per batch.
The results detail topics, actions, and responses, indicating whether they matched expected outcomes.
Agentforce & Data Cloud:
Agentforce introduces a Data Library to improve LLM responses. This library stores pre-processed chunks of knowledge articles or uploaded files (e.g., manuals, price lists, white papers), enabling Retrieval Augmented Generation (RAG).
Tips:
A Data Library cannot simultaneously support knowledge and file uploads.
Agentforce & Service Cloud:
- Einstein Article Recommendations: AI uses “Knowledge” and “Closed Cases” to display knowledge articles related to the current case. To get started, you need at least 3 knowledge articles and more than 1,000 closed cases.
- Einstein Service Replies: Generates AI-driven responses (drafts) for live chat or email, saving time for service agents.
- Einstein Work Summaries: Predicts and fills in case summaries, issues, and resolutions.
Agentforce & Sales Cloud:
- Agentforce SDR: Automates lead assignment based on predefined criteria, enabling sales representatives to efficiently manage and prioritize leads.
- Agentforce Sales Coach: Provides personalized coaching to sales representatives through role-playing sessions and pitch analysis. Designed for coaching purposes, not for managing or prioritizing leads.
- Call Explorer: With the Einstein-powered Call Explorer, users can quickly gain insights from calls, such as mentions of products or unresolved customer questions, by asking AI.
Key aspects of Agent Force:
- Autonomous operation:
Agents can perform tasks independently without needing constant human input, making decisions based on pre-defined parameters and real-time data.
- Natural language understanding:
Agents can interpret complex user requests and generate human-like responses through advanced language processing capabilities.
- Data-driven decision making:
Access to real-time information from Salesforce and other sources allows agents to tailor responses based on specific customer contexts.
- Cross-channel interaction:
Agents can engage with customers across multiple channels like web chat, email, SMS, and social media.
- Customizable workflows:
Businesses can design specific workflows and rules for their agents to follow, aligning with their brand voice and business needs.
- Handoff to human agents:
When necessary, complex issues can be seamlessly transferred to live agents for further assistance.
- Scalability:
Agentforce can handle high volumes of customer interactions consistently, providing 24/7 support.