Foodpedia

• 

  Sample output of Foodpedia

Inspiration

I say this from experience. If you're new to world cuisine, not knowing what you're ordering can lead to two things: awkward silence (because you have no idea what it is) or even worse, food envy (when your friend's dish looks way tastier). That kind of a blind date with your dinner can leave you hangry!

To avoid these kind of scenarios and educate myself about dishes that sound mysterious, I created Foodpedia

What it does

Foodpedia is an AI-powered app created with PartyRock which can provide info on food items you haven't encountered. It can tell you about that food's Origin, Appearance, Taste, Ingredients and Nutrition details. It can also tell you about comparable dishes and fun facts that you can regale your dinner companions with. Unlike with humans, you can ask the AI chatbot any question without being judged!

How I built it

The User Input, Text Generation and Chatbot widgets were configured to fetch the required food details after a lot of experimentation with the prompts. As there are now nine different Models, I tested the output of each and found Llama 2 Chat 70b better performing than Claude and the rest especially when it came to descriptions of Indian food.

Challenges I ran into

Having an image of the dish could have made the description visually appealing but almost in every case, the Image Generation widget generated pictures that were inaccurate and even hilarious.

Accomplishments that I'm proud of

It was gratifying to find a simple solution to learn more about world cuisines.

What I learned

Extensive testing can lead to better prompts and output.

What's next for Foodpedia

Now that I've built a sort of a MVP, I would like to make my app production-grade by utilizing Amazon Bedrock. I will explore ways to make the app return more accurate results cost-effectively with relevant images of eats & drinks.

Built With

• claude
• llama