This example performs a Private Set Intersection (PSI) between the customer
lists for two fictional banks -- JPB and PNB, and a dataset of Brokers.

First, synthetic data is built for both banks.  The first bank has data in
the format:
    * id
    * first_name
    * last_name
    * ssn
    * account_number

The second file of synthetic data is created in the format:
    * id
    * full_name
    * ssn
    * credit_card

The third file of synthetic data is created in the format:
    * id
    * full_name
    * license number
    * ssn


Once data has been created, all CSV files are placed on their respective
organization's Access Point, just as three different real-world organizations
would store data independently in their own data center.


Finally, PSI is performed between the Access Points using the match column
"ssn".  During the process no organization is able to see the dataset of the
others.  The calculated overlapped "ssn" values are returned as an asset
containing a simple CSV list of the overlap values.

This result asset is available in the Router index for further calculation.  In
the example script it is also downloaded to be viewed offline.
