Create Agents in Record Time with the Agentforce Implementation Guide

Build an Agent Script-powered AI agent for some of Agentforce’s top use cases with the Agentforce implementation guide.

Using a Developer Edition org, you can create a customer-facing agent that answers questions from a knowledge base and escalates for further assistance. No endless clicking, just download the PDF and go!

Watch this video to learn about the Agentforce implementation guide.

Who is Makana Medical Devices?

Makana Medical Devices is a medical device company that develops, manufactures, and distributes a wide range of diagnostic, treatment, and patient monitoring instruments. They've already integrated generative AI into their Salesforce ecosystem and are now looking to build their first agents.

Makana’s Customer-Facing Agent Helps Customers Find Information Fast

Makana wants to increase their case deflection rate, their first contact resolution, and their CSAT by launching an external-facing agent that lives on their website. This agent is powered by Agentforce Data Libraries that ingests a PDF of information and answers questions. If the agent can’t answer a question, the agent creates a case so a human can step in to help.

Follow the Agent Development Cycle

Building agents is easy when planned correctly! Using our Agent Development Cycle, you can follow Makana through the major stages of development: from ideating the first use cases to supervising the agent after launch.

Agent Development Cycle

Our guides provide helpful frameworks for you to share your plans with your key stakeholders.

Task Customer Agents
Ideate Agents
  • Identify your use case
  • Define technical requirements
  • Define agent guardrails
  • Design your agent
Build Agents
  • Set up your agent
  • Create your agent from a template
  • Customize subagents and actions
Test Agents
  • Build your testing plan
  • Manually test your agent in Agentforce Builder
  • Batch test your agent in Agentforce Testing Center
  • Troubleshoot agents
Deploy Agents
  • Create an Omni-Channel Flow
  • Launch your agent on channels
Supervise Agents
  • Post-Deployment
  • Root cause identification
  • Prepare for improvement