metamorworks - stock.adobe.com
A hyper-personalized AI platform for community finance
Using the vendor's Cortex AI platform, DeepTarget created a digital experience platform that community finance institutions can use to predict what consumers will buy.
A partnership between an AI vendor and a financial educational technology company aims to help financial institutions take the guesswork out of community financial campaigns.
On Aug. 16, CognitiveScale -- an AI vendor that specializes in hyper-personalization -- joined forces with DeepTarget, a self-described fin-ed-tech company, to introduce a new digital experience platform. Using CognitiveScale's Cortex AI technology, the DXP enables CFIs to predict whether consumers will purchase a financial product within a 30-day period, among other things.
CFIs -- credit unions, banks and funds that serve economically distressed communities -- can also use the DXP to create predictive campaigns for lending- and fee-based products for CFIs.
CognitiveScale's Cortex AI platform is aimed at data scientists, developers and business users. Developers have access to different data models and can construct models in a low-code environment. Meanwhile, data scientists can create data sources, and business users can use the platform to create AI-based marketing campaigns by establishing the goal of the business and suggesting the best ways to accomplish that goal.
Brandon PurcellAnalyst, Forrester Research
Verticalization of AI
CognitiveScale's collaboration with DeepTarget represents the recent trend of AI vendors pursuing business arrangements with other software vendors in individual markets, said Brandon Purcell, an analyst at Forrester Research.
Vendors are no longer just creating generic AI products for enterprises. Instead, a verticalization of AI is underway in which AI systems are made for specific industries, he said.
While CognitiveScale develops a predictive analytics system for enterprises, DeepTarget has deep expertise with CFIs. This partnership enables DeepTarget to create specific products addressing the challenges CFIs might face, such as customer conversion and long-term profitability.
"A lot of the machine learning vendors are missing the piece of 'OK, we've created these great AI models; now what do we do with them?'" Purcell said. "That question can only be answered in the context of the industry you're working in."
However, partnerships like these come with challenges.
One is data, Purcell said. Because CFIs tend to have a lot of consumer data, pulling that data together can be a complex, essential step before an institution even uses CognitiveScale's machine learning platform.
There can also be a problem of bias in the data because AI tends to use historical data, said Craig Le Clair, another analyst at Forrester.
"It's reflecting historical data that may not be accurate," he said.
Another challenge is making sure different departments can use the platform is a way that helps them with their clearly defined jobs. The IT and sales departments, for example, have to work together on community finance campaigns and coordinate how they are using the AI platform.
"How do we actually get salespeople who probably haven't had access to any information like these in the past to use this to become better at their jobs?" Le Clair said. "That brings up people and process types of issues that a technology vendor can't solve for."
He said privacy can also be an issue, especially because the campaigns are hyper-personalized and aimed at what a consumer may end up doing next.
"A combination of what the power of what CognitiveScale can do with the kind of domain expertise around specific marketing [and] financial services, and all of a sudden, you might have almost too personalized recommendations that will make a human think that you're being stalked or monitored," he said.
In a July interview with TechTarget Editorial, CognitiveScale's COO Shay Sabhikhi said Cortex AI enables different departments in an enterprise to work together to create a model and see how it's performing. He added that the platform takes explainability, bias and fairness into account.
"Performance is really important," Sabhikhi said. "But explainability needs to be above a certain threshold." This enables compliance officers to trust that the AI product will have the right governance and adhere to responsible AI guidelines.
DeepTarget had responsible AI and privacy in mind when choosing CognitiveScale as a partner, CEO and founder Preetha Pulusani said.
"Yes, we use consumer data, but we need to use it in a responsible way," she said.
The digital experience platform will be generally available to DeepTarget's customers in September, according to the company.