The massive amount of data collected by risk management systems often calls for a human to spend time deciphering the information and reporting on the findings. Chicago-based Narrative Science wants to automate that process.
Quill, a system the vendor designed to help compliance reporting related to anti-money laundering, identifies relevant data points and relays the information in conversational language, or what Narrative Science refers to as "natural language generation."
MasterCard is among the various companies to employ Narrative Science for this type of data reporting, as the company works within the financial services industry, government, retail, manufacturing and health-care segments.
"If there is a need to understand what is happening with that data and report on it, that is absolutely right in our wheelhouse," said Nick Beil, chief operating officer for Narrative Science.
Any area of risk management or payment data protection in which companies have to employ "armies of people and analysts looking at massive amounts of data every day" could benefit from automated narrative reports, Beil said.
While risk-management vendors seek ways to
Risk management still involves many manual tasks for companies that need written reports based on standardized data structures, said Julie Conroy, research director and fraud expert with Boston-based Aite Group. "These reports range from SAR [suspicious activity report] filings for fraud and anti-money laundering to board and management reports on key performance indicators," Conroy added.
Quill gives firms the ability to automate these tasks, Conroy said. "It frees up valuable resources for the more important business of actually fighting the fraud and working to improve customer service," she said.
Quill's automated "suspicious activity" reports have helped many banks and businesses with their compliance and fraud reporting.
"These groups have to report any detected activity, including any illegal or inappropriate movement of assets, starting with a protection model from Know Your Client information," said Kim Neuwirth, product manager at Narrative Science.
In this arena, Quill can help companies summarize and detect problems through its consistent reporting method, Neuwirth said. As such, Quill can follow a company's basic rules for employee use of corporate cards and file reports about activity that doesn't meet certain thresholds.
"It brings transparency to help companies with employee expense activity," Neuwirth said. "This is a huge amount of data, and Quill looks to interpret changes in common expense and credit card indicators."
Its reports can help companies more quickly realize when a corporate credit card has been stolen or is being abused.
"The company's system will know when employees are in New York, but what if transactions are all of the sudden coming out of California?" Neuwirth said.
Narrative Science has also created automated product descriptions for Web sites in the retail sector and leverages product data and SKUs [stock keeping unit] for automatic generation. Quill can bring that same narrative capability to fraud detection for retailers, Neuwirth said.
The business detection model sits under Quill, but Quill can detect changes in threshold and report the impact of that data change, Neuwirth added.
Ultimately, Quill is about making life easier for companies that rely on data collection — and being able to explain what that data means in a way that everyone understands.
"The theme is really around where you can leverage software to automate things," Beil said. "You want to automate so the people can actually elevate their game and focus on the more strategic tasks, and that is just a great use case for Quill."
