We wanted to satisfy a few categories at the Hampshire Hackathon 2017 and thought we could tie Artificial Intelligence together with Local Crowdsourcing to bring the easiest, cheapest way to pre-determine where you're going to buy small goods (groceries, office supplies, etc) straight to any Android device.

The categories we are submitting for are Artificial Intelligence, Life Hacks, and the Community Challenge.

What it does

In a sentence: This app plans the optimal stores to visit to purchase all the items in your shopping list and displays them in order with what items to get at that store.

The app namely uses crowdsourcing to update its database values. Users simply take a clear photo of their receipt to have the OCR determine the items purchased, their prices, the date, and the location. The values from the receipts are parsed and uploaded to update our database, which then can be queried based on the user's shopping list to suggest an optimal configuration of stores to minimize money spent and distance driven.

The value of this product is in our inherent human desire to minimize labor and maximize efficiency. The app does what no person really wants to do, but wishes was done for them.

How we built it

We built it using Android Studio in order to have an easily shippable product to the general public. The OCR we used to read the receipts is Google Vision. The Google Places API and Google Distance Matrix API were used in the backend to locate nearby stores and calculate their distances in minutes.

Challenges we ran into

The project went very smoothly up until, of course, the merging stage of all the moving parts. Our two largest downfalls were scraping data via Android Studio (on the computer there was no issue) and determining the "cheapest" path to purchase groceries.

Accomplishments that we're proud of

We worked relentlessly and are all quite new to the MLH scene. We all each learned some new specific skills, like handling new API's and creating Android apps hand-in-hand with AI. The selling strongest point of our project was the possibility of implementation, as it is extremely accomplishable with only some more time. We felt we had a strong, developed concept that would be of immediate interest to the general public.

Built With

Share this project: