In today's outsourced-dominant world, asset managers often have little control or visibility into their NAV production process. OnCorps created an AI algorithm that uses multiple data streams to predict when NAVs will be delivered late from the service provider. Developed alongside a Top 5 global asset manager, the algorithm has achieved a mean accuracy rate of 93% in just 3 months of training.
OnCorps has developed a pioneering approach to identifying and changing human behaviors. Traditional approaches to testing human behaviors are limited and infrequent. Polls and focus groups elicit what people think they will do, but not what they will actually do. Our approach, alongside our partners at Yale's Human Nature Lab, enables us to create a simulated environment that tests behavioral algorithms before they are exposed to customers. We recently applied this method to the collections function of financial institutions, where we changed behaviors to decrease time-to-pay by over 15 percent.
It’s no secret that clean, reliable data (and a lot of it) is crucial when implementing learning technologies. Within asset management operations, however, gathering ‘good data’ is met with significant roadblocks. As a result, OnCorps’ data scientists were forced to get creative when thinking about how to successfully design, train, and implement AI algorithms within fund operations.
Psychology research has proven that the human brain can handle only a certain amount of information before cognitive performance begins to decline. OnCorps’ research team thought to apply this behavioral theory to an oversight environment to test how well humans are able to process large amounts of data and identify errors.
After analyzing over a decade of NAV errors and their root causes, OnCorps’ data science team created a comprehensive, algorithmic approach to overseeing granular accounting data to identify potential NAV-impacting anomalies.
The OnCorps team was asked to build an AI platform that would reduce as many hours as possible by error checking a complex and lengthy semi-annual financial statement.