FindMyFood

A single platform to cater to all your food needs regarding restaurant and takeaway choices: FindMyFood removes the need for reviews and uses AI agents to personalise users' food preferences.

Comment

Inspiration

We've all experienced times when we thought a restaurant would be good because it was highly rated on Google Reviews, but the food turned out absolutely terrible and disappointing

What it does

FindMyFood removes the dependence on Google Reviews by using AI agents to personalise a user's food experience. The user enters whichever restaurant they want to try, the location, their food preferences (e.g. spicy, non-spicy), and dietary requirements (e.g. dairy-free, gluten-free). The AI agents use LLMs to obtain restaurant data, and the website gives the user a list of dishes which the user could order based on their options, making sure to display highly-matching dishes at the top.

How we built it

Front end: The UI was mostly designed in React typescript using the help of Figma for designs,
Backend: Typscript, OpenAi API, Python, Google places API

Challenges we ran into

One big challenge was implementing the back-end code and front-end website into one complete system, as this raised multiple errors, which took lots of time to debug and resolve. Another issue was that our initial idea (finding real Google Reviews per person and filtering which reviews to consider and which to ignore), was too complicated to actually complete, so a lot of time and effort was wasted following the original idea.

Accomplishments that we're proud of

The AI agents and website can successfully take the user's inputs, and display a variety of dishes which the user can eat and dishes which they will actually enjoy. It also displays details of the restaurant including its location (accessed via the Google Places API), ratings for dishes, descriptions of each dish and similar restaurant the user may like.

What we learned

One of the biggest learning experiences was how to connect the front-end to the back-end codes, and we learnt how to do this via Cursor and ChatGPT.

What's next for FindMyFood

To develop this further, one important feature would be enabling multiple users to enter their food preferences into FindMyFood, and the software will display dishes which each user can enjoy, all from the same restaurant. This solves the common group problem "Where should we eat?". Another feature could be the addition of a food-related social media or blog page, where food bloggers and food enthusiasts could all share their experiences and contribute to the food community, essentially making FindMyFood the one-stop solution to everything food-related.

Share this project:

Updates