AI Powered Sentiment Analysis and KPI Dashboard
Leaders:
- Joshua Soteras
- Brandon Lopez
Project Manager
- Julian Gutierrez
- Edmundo Guzman
- Francisco Guzman
- Denise Tablias
Advisor
- Huiping Guo
Project Description
QTC Leidos, officially known as Leidos QTC Health Services, is a U.S.-based healthcare organization specializing in medical examination and diagnostic services, primarily for government agencies. With a wide range of clients, gathering feedback through surveys is essential to maintaining top-tier performance and continuously improving service quality. These surveys provide valuable insights into client experiences, enabling QTC Leidos to identify areas of excellence, address concerns proactively, and implement data-driven improvements that enhance overall operational efficiency and patient satisfaction.
The project's goal is to develop and a Quality Manage System, that analyzes survey responses related to to examinee and staff appointments. A key feature is Sentiment Analysis-the process of evaluating textual responses to determine the respondent's attitude. In this context, the system classifies each client experience as positive, negative, or neutral. A LLM (Language Learning Model) Roberta utilized to analyze comments that contain sarcasm or mixed opinions, offering a more effective solution compared to traditional rule-based approaches.. The analyzed data will be used to generate data visualizations and key performance indicators (KPIs), providing actionable insights to support continuous improvement and decision-making.
A QTC Leidos employee can upload survey data in CSV format via the web-based dashboard. Upon submission, the file is parsed through the use of ASP.NET , where it is ingested, parsed, and normalized into a structured SQL Table format into Azure's database. Concurrently, sentiment analysis is performed on designated textual fields to classify responses as positive, negative, or neutral.
Project Stack
Frontend | Angular |
Backend | ASP.NET + Azure |
AI (LLM) | Roberta-Based Model (Pre-Trained) |
- Jorge Arias
- Kenneth Castro
- Darrin Du
- Javier Gonzalez
- Harshaun Khehra
- Brandon Lopez
- Walter Najera
- Andres Quezada
- Joshua Soteras
- Johny Vu