I designed and built a multi-agent AI workflow in Salesforce Agentforce to support communications, outreach, event coordination, and CRM operations. The system routes requests to specialized subagents based on user intent, allowing a single conversational interface to handle different stages of a communications workflow.

In this demo, the agent retrieves an existing Salesforce contact, develops channel-specific promotional content for Instagram, LinkedIn, and email, creates an event coordination plan, and generates a follow-up task tied directly to the contact's CRM record. The workflow uses Salesforce data when available, maintains context between requests, and uses clearly labeled placeholders rather than inventing missing information.

I also built a custom Salesforce Flow and Agentforce action that allows the agent to create follow-up tasks directly in Salesforce. The final step of the demo verifies the newly created task on the contact's record, demonstrating that the system can move beyond content generation to perform real CRM actions.

Built with: Salesforce Agentforce, Agentforce Studio, Salesforce Flow, CRM data, prompt/instruction design, multi-agent routing, and workflow automation.

This project demonstrates my approach to combining AI, communications strategy, and operational workflows to create tools that can support real-world marketing and communications teams.