Every merchant knows the loyal/key customers to his business. But he has no clue as to where they all are located, while this data lie unused in Nessie's API. To prove that useful insights can be provided for merchants from the data available, we did this application.

What it does

BScouts uses each merchant's customer details to predict/ suggest locations for business where he can be more successful if he plans to start a new business somewhere.

How we built it

We used Capital One's Nessie API to get merchant and customer data. We also populated the transactions for a set of merchants for the purpose of showing our working application. We used the data gathered to cluster customers based on the location and then rank and return top 4 revenue generating and top 4 crowded customer centers. This is done by using the zipcode of the location.

Challenges we ran into

We started off assuming we had all the data. So, when we realized we had no transactions, we had script our own data generation in a random fashion and then for few specific merchants and customers. This was time taking and almost took up one of the hackers' half day.

Accomplishments that we're proud of

Given the troubles we ran into and the minimal experience of few of us, we performed really well. We are proud of being able to complete the app successfully.

What we learned

We learned how to play with financial data and also how to use esri maps. We learnt few cool things on the way.

What's next for BScouts

BScouts uses location as the clustering parameter. We can use smarter clustering algorithm. We can also use a more complex algorithm to predict the new business, not just the location.

Built With

Share this project: