T-Data | Devpost

Home page
Trend Line Graph
Map
Topic Page

Inspiration

Making data more accessible and available to help companies understand their consumer sentiment. We wanted to create a tool that bridges the gap between consumer feedback and actionable business insights.

What it does

Our site effectively gathers data and consumer feedback by actively scraping online posts, compiling customer sentiments/feedback, and organizing it neatly within a dashboard based on consumer location. This allows businesses to visualize real-time customer opinions and spot emerging trends.

How we built it

React, JS, Python, HTML, Flask, and Tailwind.css. We combined these technologies to build a seamless platform that connects an intelligent backend with an intuitive and responsive user interface.

Challenges we ran into

We ran into challenges in connecting the frontend to the backend. It took time to properly configure the data flow and ensure the scraper’s output displayed correctly within the dashboard.

Accomplishments that we're proud of

We are proud of our functional data scraper, which has successfully scraped data from both Reddit and other social media platforms. We’re also proud of how user-friendly and efficient the dashboard turned out to be.

What we learned

The intricacies of the NVIDIA Nemotron and its application in various contexts. We used it to generate embedding inorder to form cluster reviews using K-Means. We then used two more models to automatically label those cluster and perform seniment analysis for each query.

What's next for T-Data

We aim to refine our product further and develop it into a tool that T-Mobile and other companies can utilize to gain a deeper understanding of consumer sentiment. In the future, we plan to integrate more advanced AI models for deeper sentiment analysis and predictive insights.