Demonstrates an end-to-end return and refund request handled inside the ACI Playground, Maya, the AI support agent, guides the shopper from request to resolution while the Agent Intelligence panel shows live reasoning behind every step.

• Conversational Kickoff – The shopper greets the agent and types “I want to return my order,” prompting Maya to list recent orders and ask which one to return.
• Order Selection – The shopper selects an order by ID (or via the Order Details panel), and Maya confirms the item, order number, price, and purchase date.
• Reason Capture – Maya asks for the return reason, offering quick-select options (wrong item, damaged/defective, arrived late, no longer needed, other), and the shopper selects “The item is damaged or defective.”
• Automated Resolution – Maya checks refund eligibility, confirms a physical return label is required, generates the return label, and returns a summary with the Return Order ID, refund amount, and processing timeline (7 days after the item is received).
• Behind-the-Scenes Workflow – The underlying “Return and Refund Processing” workflow is shown in the Workflows builder, visualizing the decision graph (exception checks, refund amount, refund method, store credit vs. original payment) that powers Maya’s responses — built and editable without code via Decision, Action, Gather, and Terminal nodes.
• Live Agent Intelligence – Throughout the interaction, the side panel tracks each reasoning step in real time (routing, order lookup, eligibility checks, label generation) along with per-step cost and latency.