Glide is integrated with Plaid Monitor for anti-money laundering compliance. Monitor detects if your applicants are on government watchlists, using advanced datasets that reduce false positives and increase efficiency.

Monitor screens users against a number of sanction and PEP lists. Monitor gets triggered automatically and is always run whenever a new application is submitted, regardless of role type (primary, joint, trust bene, signer, etc). Monitor screens both individuals and entities (such as trust names)

Enable the watchlists you want to screen your entities or individuals against for this program. There is no incremental cost based on the watchlists you enable - they are all included!

If you'd like to see a full list of supported watchlists see here: Plaid Monitor Watchlist Support

Date of birth filtering provides an additional data point to reduce noise in your screening program. We recommend you filter by date of birth to limit false positives of your program.

For individual-based programs, we allow you to screen your users against Politically Exposed Persons lists. As you increase in level, your likelihood of false positives significantly increases

To generate a report of all hits for audit purposes, navigate to the Plaid Dashboard and select Plaid Monitor > Reports > Create New report. Then, select the program you'd like to generate a report for, date ranges, HIT types and select generate.

The name sensitivity you set has a direct impact on the type of hits you get. The different levels do not turn on or off any of our fuzzy matching features, but they limit results based on a score we compute internally. We have tuned the sensitivity levels and strongly recommend you try Balanced and adjust from there.

We have many algorithms that take into account things like: normalization (lower case vs. upper case, stop words like Mr., Dr., etc), phonetic similarity, complete mismatches on parts of the name, reversed orders, initialisms, gender adjustments, and many more.

Each type of matching phenomenon will have a different impact on the overall score, based on the importance of the discrepancy, the length and composition of the name being verified, and other variables like language. For example, “Mr. John Doe” vs. “John Doe” will receive a negligible penalty, whereas “John Paul Doe” vs “John Doe” will receive a more significant penalty. This means there isn’t a hard differentiation between the different levels, they just set different thresholds based on the scores that get computed internally.

The ranges are currently [0.7, 0.8) for Coarse, [0.8, 0.9) for Balanced, [0.9, 1.0) for Strict, exactly 1.0 for Exact. These scores can be equated to scores similar to what you may have used with other solutions:

That said, setting the scores at the same level between different solutions will not necessarily yield the same results, as the underlying fuzzy matching algorithms would need to be the same across the board. We believe our algorithms are much more sophisticated than many competing solutions which results in fewer false positives, while ensuring that no relevant hits are missed. Thus the numbers above can be used as a reference point, but we strongly suggest starting at a Balanced sensitivity level, and adjusting up or down from there.