Orchestrating an automated multi-agent workflow
Our second case study showcases how AI agents can be used within workflow automation. Our company, PaperCo, wants to periodically send personalized follow-up emails to its customers, catching up on their interests and subtly promoting a new product offering. Instead of manually researching each customer and crafting an email, we can build an agent system to do this automatically.
This workflow will involve two agents working in sequence:
- Customer research agent: This agent gathers information about a customer – including basic details from a database, recent conversation transcripts (to recall their interests or any personal information they mentioned), and even current news related to those interests (via a web search)
- Email creation agent: This agent takes the compiled information from the research agent and generates a short, personalized email for the customer, while also mentioning the new product offer...
