GroupGrub

Inspiration

One of our team members was sharing biryani with a friend in a small town when he mentioned his love for trying new cultural cuisines. He noted that it was especially difficult to find places that could fulfill all of his friend group’s cuisine dreams. When you live in a small town and have an indecisive group of friends, you learn to cherish all of the unique cuisine you stumble across that you and all of your friends wished you had found sooner. In bigger cities, the quest for trying new foods is an even bigger challenge. There are so many fast food places and local businesses that it’s hard to keep track of the new restaurant on the block you’ve been meaning to try. If only there was a way to explore good food and all the smaller, more obscure restaurants that have delicious home-cooked meals. _ Oh wait, there is… _

What GroupGrub does

GroupGrub is a social foodie app that curates top restaurant picks in your area based on you and your friends’ personal food preferences. Simply create a new group session, scan the QR code, and update your cuisine preferences and what you feel like eating that day. GroupGrub is designed to take your group’s preferred food data and find the best reviewed restaurants near you. It is specifically trained on AI modeling to curate the perfect restaurant list so you and your friends can enjoy good cuisine without the indecisiveness. The app is trained to match everyone’s unique dietary restrictions, cuisine desires, budget, and location to ensure recommendations that best suits the group.

The Building Process

We used React Native and Expo to create the mobile app. We used Firebase Firestore to store the data from the React Native end. Using the data stored on the Firestore, we call our recommendation algorithm which is done using Vespa using text embeddings. This also allows us to get a relevance score for all of our text matching that we perform.

Challenges we ran into

During our project's evolution, we encountered a series of unforeseen bugs, including some unique ones stemming from disparities between Unix and Windows environments, notably in spacing conventions. These differences often led to checksum mismatches in our certificates, triggering errors that demanded meticulous troubleshooting. Furthermore, integrating new tools and products, such as Vespa, presented a steep learning curve. Our initiation with Vespa posed significant challenges, particularly in its setup and comprehension of its features. Nonetheless, each hurdle served as a valuable learning opportunity, enriching our expertise in software development and integration processes while offering insights into the complexities of the technologies at play.

Accomplishments that we're proud of

We created and implemented a QR code scanning system, which allows users of the app to scan their friend’s group sessions and curate restaurants specifically to multiple friends’ food preferences. We used Vespa’s AI tooling to take restaurant reviews and categories of foods and translate it to curating a perfect restaurant feed for multiple data preferences. We also used location pinpointing to find restaurants that matched the group’s preferences but leveraged Vespa’s AI to prioritize catering to food preferences even if they were a little further away than other restaurants. We’re proud of our uniform logo and color scheme as they apply to the marketing world, since we are marketing our product for social foodies and other sponsors who want to give us an opportunity to continue building features of the app. We hope to brand GroupGrub as an app that can truly change how the world experiences cuisine and shares experiences with food together.

What we learned

This app development experience has been a learning experience for every one of our team members. Members of our team were exposed to React Native for the first time, and developing a mobile application was a first for a lot of us. The shift from Web Development to Mobile App Development was refreshing in how it taught a lot of us different debugging techniques, since we had to avoid using web development tools. For the designing side of the team, they learned to design an entire brand and uniform mobile app from the ground up. This was encouraging as it taught the team how to start their own brand or company someday and that the team could, in fact, do it as we did during this event. We learned to not only work with one another but work with mentors from all over the industry as well as work with those who created the software itself.

What's next for GroupGrub

We’re hoping to scale GroupGrub to a full-grown app that people can download to change the way people find food together. GroupGrub has the potential to be a widely used platform where people can share and find other foodie lovers, curate their restaurant experiences, and use a variety of other features implemented in the future, such as a receipt scanner built in the app that would scan a receipt and easily split the bill between friends using third party applications like Venmo.

With GroupGrub’s implementation of AI agents, machine learning, and many opportunities to support other cool features, GroupGrub has the potential to grow into an app shared by all the people we love to grab food with. GroupGrub takes care of all the little things so you can focus on the bigger things.

Stay in the moment.

Use GroupGrub :)