Rest-Review

Ever want a TLDR of a restaurant? Rest-Review returns recommended dishes and a summary of the atmosphere based on previous restaurant reviewers!

Comment

Inspiration

We wanted an easier way to find recommended dishes at a restaurant as well as a concise summary of the general reviews.

What it does

Rest-Review returns a concise summary of the general sentiment of the restaurant as well as a list of recommended dishes. No need to read multiple reviews, just read Rest-Review and the whole picture!

How we built it

We used Google Place API to get the location from user input as well as the reviews for the given location. We then fed the reviews in OpenAI gpt-3.5-turbo. This returned the recommended dishes and general summary that we displayed on the front-end with React.

Challenges we ran into

Accessing and configuring each API to communicate with each other, the time crunch, using the Google Maps API to properly display the location based on user input and get reviews, and creating prompts for gpt to return correct responses.

Accomplishments that we're proud of

Being able to create a fluid front-end that properly returned a good summary of dishes and atmosphere.

What we learned

Accessing OpenAI api's for ml models to accept and respond dynamically. Learned the React framework to create the front-end.

What's next for Rest-Review

Add the ability to save locations and add a parameter to recommend specific dishes, perhaps dietary restrictions.

Built With

Share this project:

Updates