Inspiration
We understand that for many people it is hard to find time or cannot leave the house to get groceries to make healthy home-cooked meals. We wanted to use the power of DAIN to help these individuals find recipes and order things online.
What it does
The add-on lets users chat with a bot and find recipe suggestions with detailed ingredient lists. It also has the functionality of identifying nearby stores that you can order groceries from.
How we built it
We used DAIN's base and added toolConfigs which allows the chatbot to call APIs. We used theMealDB to access recipes and Google Maps API to find nearby stores.
Challenges we ran into
Getting access to the APIs we wanted to fully create our vision was hard. We originally planned to allow users to order groceries directly through the chatbot, but we were not able to get authorization from sites like Amazon Fresh and Walmart, which would have allowed us to perform these actions. It was also difficult getting the chatbot to output what we wanted it to since if any of the tools failed, the entire functionality of our add-on would cease.
Accomplishments that we're proud of
We are proud of our ability to get the AI working with the recipe suggestions and store locator.
What we learned
We learned how difficult it is to access APIs online and also how hard it is to work with AI chatbots. We learned how to organize our idea of an app into smaller actions that the bot could understand and make use of.
What's next for FlavorFetch
While we did run out of time, we wanted to incorporate a script that orders the ingredients for users. For example, the script would take the ingredients and open a delivery service app such as Instacart or AmazonFresh to place an order to the user's house.
Built With
- dain
- google-maps
- node.js
- themealdb-api
- typescript
Log in or sign up for Devpost to join the conversation.