Run     User Time       System Time     Total CPU Time     Wall Time   Total Rows      Rows Retrieved
1       .42             .14             .56                1:36.6      118,000         118,000
2       .38             .12             .50                5:42.9      25,000,000      26,000

Hypothesis:
The server is taking a long time to respond to the request, which is causing the long response time. Both scripts spend
very little time running, which means they're probably spending more time waiting for data from the server. I suspect
that the R script may be faster because they're possibly running on the same machine as the server, so the network
request stays local to the server, whereas the Python script is being run externally from the server, so the network is
necessarily slower.

Steps to reproduce:
1. Create some tables to emulate the query provided by the customer
2. Populate these tables with a similar size of data (25M total rows, that can be filtered to 26k rows)
3. Run the script locally with profiler.


API time: 38.3407 seconds
Pandas time: 1.5355 seconds
Total time: 39.8762 seconds