AgentHack submission type
Enterprise Agents
Name
edip
How many agents do you use
One agent
Industry category in which use case would best fit in (Select up to 2 industries)
Customer service
Complexity level
Intermediate
Summary (abstract)
Customer Complaint Sentiment Assistant
This agent analyzes customer complaints and determines their emotional tone, issue type, and urgency. It then decides whether to auto-respond, escalate, or log the case. Built using UiPath Agent Builder + Studio Web, this MVP offers a structured, smart triage for customer feedback handling.
You can view the full demo presentation including workflow steps and architecture via the following link:
Demo Presentation (Google Slides)
Detailed problem statement
Most RPA workflows process customer complaints as structured data, ignoring emotional tone. As a result, angry or urgent complaints receive the same response as routine ones.
This causes delays, missed escalations, and poor customer satisfaction. There is a need for an intelligent automation that can evaluate the emotional intensity of customer messages and act accordingly.
Detailed solution
We built an agent that reads customer complaints (via Excel in MVP, API/portal in future), analyzes tone using a system prompt, and returns structured results including:
Tone (e.g., Angry, Neutral)
Category (e.g., Payment Issue)
Priority (e.g., Urgent)
If the tone is critical (e.g., Angry), it generates:
An Urgent Help Case summary
A Company Reply in a polite tone
It also checks previous complaint history using a ReadHistory tool. Final output is formatted in JSON and ready for downstream tools (e.g., logging or RPA queue).
Future plans include a UiPath Apps interface, Maestro orchestration, and LLM-based enhancements.
Demo Video
Expected impact of this automation
Time Saved: Faster classification reduces human triage effort
improved Accuracy: Emotional tone guides smarter decisions
Reduced Escalation Delays: Urgent complaints are flagged instantly
Automation Readiness: Output can directly feed RPA queue
ROI Potential: Reduced churn through improved customer handling
Extensible Design: Can integrate with LLMs, UIs, queues, or APIs
UiPath products used (select up to 4 items)
UiPath AI Center
UiPath Agent Builder
UiPath Apps
UiPath Studio Web
Automation Applications
Customer Service (CRM), Complaint Handling, Contact Centers
Integration with external technologies
OpenAI / LLMs (via prompt-based analysis) Excel (temporary input source in MVP) REST APIs (planned)
Agentic solution architecture (file size up to 4 MB)
Sample inputs and outputs for solution execution
10. Sample Inputs and Outputs
Inputs:
Complaint text (from Excel or future API)
Optional: customer name or ID
Outputs:
Tone, Category, Priority, Explanation
If needed: UrgentHelpCase and CompanyReply
JSON result formatted for activity history logging
Other resources
n