Inspiration

As hungry students, we are easily swayed by food. After hearing the Radish presentation during the opening ceremonies, we were inspired by their mission to connect people with food. We also like good, cheap food, and so we were inspired to create something that would suggest popular and inexpensive meals at restaurants in the area. We were inundated with ideas, but were forced to be more realistic, given our limited (nonexistent) prior hackathon experience.

What it does

Our website allows the user to input in a search bar the food that they are looking for. Then, this information is sent to the backend, where a dictionary with hundred(s) of restaurants and their menus are searched. If the "boujee" checkbox is not ticked, the application will find the cheapest (up to three) options of food in all the menus of the restaurants. Then, the restaurant names, menu item, and price will be displayed. If "boujee" is selected, the app works this same way, this time displaying the most expensive items.

How we built it

We began by building the backend. First, we wrote Python code to manipulate and sift through a handwritten dictionary of restaurants, to find the cheapest/most expensive options, by analyzing the prices associated to menu items and keeping track of all the variables. Then, we began development of the Flask website, learning html along the way. Soon enough, we were working on connecting the back end to the front end. We also used co:here to generate via AI a much bigger dictionary of 100 restaurants, although the size could easily be enlarged. We also spent some time working on the aesthetics of the website, including learning how to upload an image, learning css, drawing a background.

Challenges we ran into

There was a steep learning curve for all of us at the beginning. Since we've never been to a hackathon before, there was a lot we had to learn about web and software development. We for sure made a mentor, David, go a bit crazy with our questions, but he helped us a lot. We were trying to build a website through Velo for over 5 hours, until we switched to Flask. We were also trying to collect data from the real world, which is definitely a future possibility, but it took a long time to realize it wasn't realistic for us to try to collect data about restaurants from UberEats or from Radish.

Accomplishments that we're proud of

We're really proud of using AI to generate a huge list of restaurants. We provided the AI with a few examples, and then used the random module to select a random cuisine, the AI was able to generate realistic and creative outputs. This was a really cool thing to learn, and we had been trying to implement it all along, so it was very satisfying for it to come together in the end. We're also proud of persevering and learning how to write and design a website in Flask, because it wasn't intuitive for us and we spend a lot of time learning how to do that. We even managed to learn how to upload images for display.

What we learned

As mentioned previously, we learned so much about web development because we were coming from ground zero, and only knew Python. We didn't know html at all, so we learned a lot about that through YouTube tutorials. We also learned how to use AI and that was very cool.

What's next for CheapYums

As we mentioned previously, we had to reign in our ambitions during the brainstorming process. Originally, we had hoped to provide a very interactive interface, through which users could select their mood, their comfort foods, and more. In return, the platform would build a food profile for the user, and would suggest users with cheap restaurants nearby, as well as filtering through Google reviews for a suggested restaurant. The platform would find the best and worst dishes as rated by customers, and would suggest to them the best one, and to their enemy (or blocked contacts) the worst dishes :). As well, ideally, we would implement a box for the user to customize how many suggestions they want to receive. Needless to say, we have many more ideas about how to bring this project even further to life!

Built With

Share this project:

Updates