Transaction Network 1: Join Users, Accounts, and Money Flows

The challenge

Follow the connections between users, accounts, and transactions. Build a network you can explore, find its busiest accounts, and create a simple rule for choosing activity to review.

Start here

Open the SharePoint folder linked above. users.csv, accounts.csv, and transactions.csv are the primary inputs. transactions.xlsx is another tabular export. generation_report.json and summary.txt are aggregate reference reports. chord_diagram_data.json is a roughly 691 MB dense matrix export; inspect its opening structure instead of loading it to begin the exercise.

There is no raw PDF in this folder. The useful conversion is from linked flat files to a typed investigation table and a compact edge list.

Try it

1. Join transaction endpoints to their account owners and validate the joins without multiplying rows.

2. Create a sparse network representation and calculate account activity or counterparty features.

3. Reconcile counts with the generation report, including the meaning of fraud_rate and risk_flags.

4. Summarize amounts by currency and design an explainable review ranking without leaking answer labels.

What to build

Build typed users, accounts, and transactions tables; an edge CSV or JSON Lines file; a join-quality report; and a short ranked example with reasons and a stated evaluation unit.