Smart Support Triager

• 

  Exception Path: High-risk intents are automatically routed to a human via UiPath Action Center.

• 

  Happy Path: Simple requests are resolved instantly with an automated RPA database lookup.

Inspiration

Customer support teams spend countless hours manually reading and triaging emails. Simple requests (like "Where is my order?") take time away from complex issues (like angry customers demanding refunds). We wanted to build an agentic solution that acts as a highly intelligent "first line of defense," utilizing AI to understand customer intent and UiPath Maestro to orchestrate the downstream actions.

What it does

The Smart Support Triager sits between the customer's inbox and the support team. When an email arrives, it triggers our workflow:

1. AI Classification: An AI agent reads the email to determine the underlying intent and the customer's sentiment.
2. BPMN Logic Gateway: The UiPath Maestro flow evaluates the AI's classification.
3. Automated Resolution: If the email is a simple status check, Maestro triggers an RPA task to draft a reply.
4. Human Escalation: If the customer is angry or requesting a refund, Maestro automatically routes the ticket to the UiPath Action Center for immediate human review. ## How we built it Since we had a strict timeline, we rapidly prototyped the agentic logic using Python and created an interactive frontend to demonstrate the flow.
5. The Brain: We designed the AI classification schema to reliably output structured JSON.
6. The Orchestration: We mapped out the Maestro BPMN logic to handle conditional routing based on the AI's structured output.
7. The UI: We built a custom, interactive web interface using HTML, CSS (Glassmorphism), and Vanilla JavaScript to visually simulate the exact steps Maestro takes. ## Challenges we ran into The biggest challenge was designing a system that doesn't just "answer" questions, but actually knows when to hand off to a human. Determining the exact routing logic for the BPMN gateway required careful thought about which sentiments and intents require human empathy. ## Accomplishments that we're proud of We are incredibly proud of completing a fully functional, visually stunning prototype of a complex enterprise workflow. The custom UI clearly demonstrates the power of combining AI Agents with structured BPMN orchestration. ## What we learned We learned that combining AI with structured BPMN workflows creates a much more reliable and enterprise-ready solution than relying on AI alone. ## What's next for Smart Support Triager In the future, we plan to connect this directly to a live email inbox using UiPath Integration Service, integrate it with a real CRM, and expand the AI Agent to handle dozens of different intent classifications.

Built With

• css
• html
• python
• uipath