Oversight is an Atlanta-based software (as a service) company founded back in 2003 to help organizations monitor spending in an effort to identify errors, waste, misuse, and fraud in the grey area of enterprise spend. As every recovery firm will tell you, the average organization will overspend by 1% to 3% as a result of over billings, duplicate billings, unnecessary spend on superfluous demand, maverick spend, and even fraud. (And they make their living recovering a portion of that, typically a third, and then charging you 33% of the recovery as their fee. Sounds small, but 1/3 of 1/3 of 3% of spend is 0.33% of spend, and if the organization spends 100 Million, they get 330,000 for an effort that can be largely automated and, even worse, be avoided with proper up-front spend monitoring.)

For example, if all invoices are compared to invoices and goods receipts before payments are authorized, this can prevent overpayments. Duplicate billings can be identified in the same way (and duplicate payments prevented). Potential fraud can be identified by forcing all invoices from unknown suppliers, for unknown products, or for unexpected amounts to be manually reviewed. (This can’t prevent in-house fraud, where a buyer pays a fake invoice to a fake company controlled by a relative, or a co-conspirator, but it can prevent external fraud.) Unnecessary spend on superfluous demand will require up front requisition control, as will maverick spend, but at least there will be no overspend or duplicate spend that can be unrecoverable once the contract with the supplier expires.

Oversight is unique in that it is not so much a software platform but an insights platform. Employing a team of data scientists focussed on identifying new algorithms and techniques for fraud detection, Oversight uses their in-depth knowledge of fraud to build solutions that will help the clients identify potential cases of fraud that they could never hope to identify on their own. The best most companies can do is sample based audits and spot checks which are unlikely to identify much fraud as these will generally only be on a few percentage of invoices or transactions, and most employees who have been getting away with fraud for a while will not be doing anything obvious, and the fraud will not be detected without correlations across documents and systems. That’s where Oversight comes in.

The Oversight solution is a web-based software solution for automatic spend analysis and identification of high-risk or potentially fraudulent transactions that comprehensively analyzes T&E, purchase card, and accounts payable spend using a suite of statistical, clustering, data mining, break point, rule-based, evidentiary reasoning, and machine learning algorithms that look for discrepancies, suspicious patterns, known fraud, and risk indicators to identify those transactions that need to be manually reviewed. The dashboard-driven, or work-bench driven, interface allows an analyst to drill into suspicious transactions by country, organizational unit, risk level, or exception type and can be configured to show the analyst only those exceptions assigned to her, or her team, or every unresolved exception in the system.

When a user drills in by exception type, she sees an overview of the overall risks by country and can drill into suppliers to see the specific exceptions. When a user drills in by country, she can see the overall risk by supplier and then by exception. In other words, she can drill into at-risk transactions using country, organizational unit, supplier, and at-risk type in any manner they please.

Or, they can look for exceptions by process. Right now, Oversight supports the identification of at-risk transactions in the travel & expense, procure to pay, and purchase card processes and has recently added support for FCPA, Anti-Bribery, and Corruption Risk — including the identification of known politically exposed parties.

Plus, the platform not only integrates with all of the big supplier and financial data providers — such as Dunn & Bradstreet, Bureau van Dijk, and CreditSafe — but also integrates with providers of risk indicator data such as Ecovadis and Sedex Global. Plus, they maintain their own databases of known politically connected parties, gentlemen’s clubs, denied parties, and other parties that an organization typically should not be allocating funds to. This last capability is quite important … just ask American Express which once received a 241K strip club bill authorized by the CEO. (Source)

Since fraud attempts differ by country, and collusion is hard to detect with a standard m-way match invoice processing platform, Oversight brings a powerful offering to the expense management space. It’s a platform worth checking out. For a deeper dive into the platform, check out the recent coverage by the doctor and the prophet over on Spend Matters Pro [membership required]. (Part I is up with Parts II and III coming within a week.)