Written by Aidar Orunkhanov
It’s Friday afternoon. You’ve left your cape and utility belt in the office, and are heading home for a weekend of socializing. You will keep friends and families on the edge of their seats with stories from the front-ish lines in the war against terrorism, drug trafficking, money laundering, corruption, and fraud.
You are not just one of your bank’s top crime fighters, you are the most interesting compliance officer in the world.
Tamr does a lot to help compliance and investigative teams like yours–providing software to your chief data officer and IT department to support KYC, AML and other compliance and risk initiatives on a global scale.
But today, we’re supporting you with conversation fodder for the weekend. Party goers will be blown away when they hear you tell them about…(drumroll)…Master Data Management.
Maybe you’ve heard your IT team talk about MDM, maybe not, but you’ll definitely want to lay this on people.
“With all the millions of retail banking customers my bank has around the world…did you ever wonder how we find the money launderers? It’s not easy…they’re clever and highly motivated to conduct banking transactions that fly under our radar.”
“We store information about customers in hundreds of IT programs…in different departments, in different countries, in different languages. The crooks think it’s hard for us to keep track of what they do, with record keeping spread so far and wide, but you know what we have?”
“Master data management. Boom!”
MDM software takes information about customers from the hundreds of IT systems in the bank, and puts it in a central place. Then it resolves duplicate records and such until you have one master record for each customer.
It’s way easier said than done because each IT system can have different information about each customer. Or store it in a different format, like US vs. European dates. Plus, there are data entry inconsistencies and errors. So the MDM process is pretty complex, but basically works as follows:
- First, it combines data into a central database (or sometimes called a “data lake”)
- Next, it conforms that data into a standard format
- Finally, it clusters matching or duplicate records of the same person or business together to form the master record
The whole process is exactly like if 100 co-workers, each in different countries, sent you a spreadsheet with 50,000 customers in it…and you had to combine them into a single sheet and deduplicate it. Tedious!
Once the records are mastered, then you can use them to get a unified view of the business each customer conducts and segment them into different risk categories based on who they are and what kind of business they conduct (Customer Due Diligence).
And now it becomes clear that Uncle Walter, a teacher and a holder of a checking account with very low activity, is actually trying to hide his identity as a beneficial owner of dozens of shell companies that use their accounts to launder funds acquired from suspicious activity.
This is a simple definition of MDM… Like we said, it’s harder than it sounds. Some companies, like multinationals or those that are more diversified, have bigger MDM problems than others. And depending on which MDM system you use, it can be a harder or easier, or a more costly or less expensive problem to solve.
But that detail will have to wait for another party.
In the meantime, if you need other interesting topics of conversation, we recommend:
Have a good week, and thanks for keeping the world safe.