Inspiration
As a team made up mostly of business majors, we were looking for a challenge that would intersect business and tech. We wanted to show that even without being pure engineers, a team with strong business intuition can design a system that turns customer feedback into actionable intelligence.
What it does
Repuvance starts by leveraging APIs to gather data from sources such as Reddit, Google Maps, YouTube, DownDetector, etc. Using Aspect-Based Sentiment Analysis with weighted scoring, it evaluates feedback by interaction quality and source credibility to calculate dynamic Net Promoter Scores (NPS). Insights can be filtered by category, source, or timeframe, allowing businesses to narrow down their leverage points. Visualization tools such as trend charts and location based heatmaps are powerful tools that provide insights into where customer satisfaction lacks or out performs. Additionally, AI integration delivers actionable insights classified as “Opportunities to Amplify” or “Risks to Address”. Tasks are also generated into TMF Tickets with key details (category, severity, task name, description, ID, date, and assignee) for seamless integration with project management tools like Jira. As a finishing touch, a competitor analysis module compares weighted NPS and sentiment trends, highlighting competitive pain points and opportunities, enabling businesses to gain a measurable advantage.
How we built it
We built Repuvance with a combination of Python, AWS, and React. We designed AI agents with AWS who played a huge role in organizing our data, as well as delivering the AI analysis. Our data visualizations were created with Tableau and later integrated into our User Interface, which was created with the help of AI tools.
Challenges we ran into
Bringing together so many data sources was tough. Every platform structures feedback differently. Integrating TM Forum APIs was also a learning curve and something that none of us had done before. Lastly, of course, turning the backend code into a presentable User Interface was a great challenge.
Accomplishments that we're proud of
We are extremely proud to have been able to successfully retrieve data from various online sources, allowing us to work with real data and give meaningful insights. Building a product that merges data engineering, AI, and business knowledge into one cohesive experience was also a huge win for us.
What we learned
Three out of the Four of us are business majors, making this a roller coaster of a project. We learned that no matter your background, if you commit and work your butt off, you can achieve something great. With a great team mentor, us business majors were able to learn to read and write code, and put together a great product.
What's next for Repuvance
Repuvance plans to lead the next generation of customer intelligence through automated systems that merge private data and public discourse into actionable, real-time insights.
Log in or sign up for Devpost to join the conversation.