Status : Verified
| Personal Name | Quiban, Gann Sebastian B. |
|---|---|
| Resource Title | Development of an AI-driven DSAT VOC scrubbing solution for contact centers |
| Date Issued | 11 June 2025 |
| Abstract | In the customer service industry, customer feedback provides rich insights that can benefit BPO companies seeking to continuously improve the processes and services for end customers. However, these insights are often underutilized due to the lack of efficient analysis solutions. Traditional methods are manual, time-consuming, and prone to human error, resulting in an average of 480 seconds of scrubbing time per comment. This capstone aims to develop an AI-powered solution to automate and enhance the analysis of each DSAT comment received and classify it into the DSAT attributes present in the customer’s comment using Microsoft Power Platform. The solution incorporates features such as sentiment analysis, and multiple custom RAG-based prompts that power the AI automated classification process. The testing results show a 96.85% reduction in processing time from 480 seconds to 15.1 seconds per comment, significantly reducing the scrubbing time and improving the leader’s accessibility to the insights. This enables faster execution of action plans to address the opportunities identified by the solution. The AI-driven DSAT VOC scrubbing solution promotes scrubbing at scale, where BPO companies can continuously harness unstructured feedback data to drive service delivery, customer satisfaction, and innovation. |
| Degree Course | Master of Technology Management |
| Language | English |
| Keyword | RAG; Automation; VOC mining; LLM; VOC; Large Language Model; Voice of the Customer; Business Process Outsourcing; GPT; Low-Code/ No-Code; Text Analytics; Multi-Label Classification; Data Analysis; Contact Center; Call Center; BPO; Artificial Intelligence; AI; Retrieval Augmented Generation; DSAT |
| Material Type | Thesis/Dissertation |
Preliminary Pages
111.21 Kb
Category : C - Confidential information of a third-party is embedded.
Access Permission : Limited Access