As a business owner, director of a club, or anyone who provides a good or service, it can be difficult to gauge customer response without sifting through hundreds of reviews manually. ReVal is designed to streamline the customer analytics experience and provide powerful insights into your product or service, made beautiful.

What it does

Drag your exported reviews/feedback data into ReVal and gain access to advanced customer response tools in a sleek, interactive user interface. We do natural language processing in Python to determine review sentiments and provide a way for users to know where their product is succeeding or lacking.

How we built it

The frontend is made with React, TailwindCSS, and Next.js, hosted on Vercel. The backend utilizes the natural language toolkit, the summa library, and pandas. It also has an API made with Flask and SQLalchemy.

Challenges we ran into

It was our first time doing natural language processing so it was a struggle to figure out how to parse though the reviews and gain insights based on the language used. Creating the API and utilizing a database was also challenging. A challenge on the front end was visualizing the data, since none of us had any experience making charts in React. Another challenge was designing the front-end to be sleek and intuitive.

Accomplishments that we're proud of

We are proud of creating our first data visualization web app.

What we learned

We learned a lot about NLP and connecting front-end to back-end via APIs. It was also interesting to learn how different product owners think when it comes to reviews/feedback on their product. We learned how important it is to draw insights from feedback using technology. We've all seen google feedback forms with 15-20 responses, but working with datasets containing thousands of reviews showed us the scale of this global challenge!

What's next for ReVal

The next step is for ReVal to be able to allow users to sort by more categories and provide more insights such as sarcasm detection and using different NLP models for different geographical locations. Different people around the world communicate in many different ways, so it's not viable to restrict ourselves to one general model. We want to deploy this app to make it usable for clubs at our university. Being able to visualize this kind of data is extremely powerful, and we want to share it with others!

Built With

Share this project: