I located a csv file with timestamped data on daily expenses. The file included several columns like category, subcategory of expenses, as well as the amount of the expense. This is the original dataset https://www.kaggle.com/datasets/prasad22/daily-transactions-dataset

I designed a producer and consumer to send and receive alerts on expenses over $5000 using RabbitMQ. Becuase I was only interested in this one alert, I did not need to use multiple queues.

Some entries included dates, but not a timestamp. For the purposes of this project about streaming data, I excluded these entries through the python code. To simulate streaming, I added a sleep time, but because I wanted to see if it was working quickly, I set this to just 1.

I had not done the optional assignment of sending an alert to email, so I tackled that this time around. Suggestions from classmates helped me figure out how to use .gitignore and .env to store my password and from_email. I also worked around gmail errors by creating a two step authentification app password. I had to keep adjsuting when the repo was pushed to github, but I finally got it to leave the .env out.

The result is a simple producer and consumer app similar to existing banking apps that sends an email alert whenever it finds a daily expense of 5000 or more.

Here is the expense alerts queue running in RabbitMQ

Here is the producer and consumer running is separate terminals indicating email alerts are being sent

Here are the email alerts in my inbox