As new grads, our team members are moving from small towns to big cities. In order to help us figure out our new financial situation, we developed Budget.io. With Budget.io, we are able to calculate personalized budgets for our new lifestyles. Furthermore, we are able to visualize data on neighborhoods, nearby restaurants, and safety ratings to choose an ideal place to live.
What it does
Budget.io uses machine learning (linear regression) models trained on data from the Bureau of Labor Statistics to predict a good budget given previous and current salaries, and the budget in the old city. We have also scraped various sources of neighborhood related data to add to our databases and Microstrategy visualizations. We are able to look at crime heatmaps, housing prices, pictures of neighborhoods, nearby restaurant and their information, etc. in an interactive map visualization.
How I built it
We first scoured the web for relevant datasets. We were able to find sets that covered bits and pieces of what we want, but it was up to us to clean and merge various datasets into something actually helpful for our project. We also scraped data through various API's such as Zomato's and worked through many different database schemas to use with our Django web app and Microstrategy visualization. We registered a custom domain name and deployed our web application to it through Google Cloud's AppEngine. Finally, we connected Microstrategy's visualization tools like dossiers to our web app and data.
Challenges I ran into
The Microstrategy environment went down around 1AM Sunday, and did not go back up until 8:33AM. We have been in limbo for a core functionality of our project during the time between. There were a lot of challenges with using Microstrategy's API.
Accomplishments that I'm proud of
We were able to scrape, clean, and synthesize massive amounts of diverse data. We were also able to get our project working.
What I learned
We learned a lot about using Microstrategy's tools as well as collected and processing data from various sources.
What's next for Budget IO
More data, more visualizations!
Our source code exceeds the 35MB limit on devpost. You can access our github repo here: https://github.com/DavidCThames/budgetio