Inspiration
In homes where they are rich enough to always have a well-stocked fridge, i noticed that they always have something going off, and they are buying the food they already have at home but forget to buy the ones that are actually running low. So they end up with an incomplete shopping trip, and extra food that they forget to use. This app i built is an attempt to solve that problem.
What it does
This app
How we built it
I first prepared the prompt using ChatGPT, after that i pasted the desired prompt into MeDo
Challenges we ran into
the refinements that i needed to do also created surprising discoveries. for one, when i directly asked for a fix the Medo agent would fail to make the fix, but i cases when i did not ask for the fix, but wanting to understand why the bug occured, surprisingly the MeDo agent made the real fix. Frustrating because i wasted credits
Accomplishments that we're proud of
I am proud that this app was mostly built, like more that half to near completion with just one prompt, everything else was basically refinements.
What we learned
I learned that to be able to get the agent to behave as you expect it to, is only possible by having enough credits to burn to go through the process of trying and failing to learn the way the agent functions.
What's next for PlatterPantry
I recognise that the process of logging food items into the app will still be a hassle even with the use of STT and multi-modal parsing. So a solution to this is to make it competitive by letting children to the loggin and getting rewards for their effort.
Built With
- css
- llm-plugin
- medo
- next.js
- react
- stt-plugin
- supabase
- tailwind
- typescript
Log in or sign up for Devpost to join the conversation.