Financial data network Plaid has launched transaction categories enhanced by artificial intelligence (AI) that it said will provide the accuracy and insight needed for the next generation of financial services.
“Financial experiences are rapidly shifting toward personalized, predictive interactions powered by AI,” the company said in a Wednesday (Dec. 3) blog post. “Delivering those experiences requires highly accurate transaction data, not just raw descriptors.”
Plaid’s upgraded transaction categorization model delivers up to 10% higher accuracy on primary categories and 20% higher accuracy on detail subcategories, according to the post.
The model delivers more reliable categories by using AI-assisted label generation and targeted human review, the post said.
It also introduces more than a dozen new subcategories for income, repayments, disbursements, bank fees and transfers. For example, new subcategories for income include gig economy, child support, rental income, long-term disability, military and retirement pension, per the post.
These refinements provide Plaid’s customers with more detailed data that they can use to design products. For example, with these new categories, digital finance providers can better understand users’ earnings, earned wage access providers can forecast payouts and adjust limits, and personal financial management apps can deliver more accurate tax estimates and personalized insights, according to the post.
“As AI reshapes financial services, high-quality data becomes essential,” the post said. “We’re continually improving Transactions to deliver more intelligent enrichment and predictive capabilities.”
PYMNTS reported in June that Plaid launched 12 years ago primarily as a data connectivity layer, allowing consumers to link bank accounts to FinTech applications, but is now evolving into a sophisticated platform focused on enhancing financial analytics and enabling new digital experiences.
Interviewed at the time by PYMNTS CEO Karen Webster, Plaid CEO Zach Perret highlighted the company’s shift toward leveraging the vast amounts of data flowing through its platform to solve complex problems like financial fraud, improve credit scoring and enhance payments analytics.
Looking five years ahead, Perret said he wants Plaid to be known as “the analytics platform for financial services or the data platform for financial services.”
Plaid and FICO said in November that they launched a credit score that uses traditional data compiled by FICO and real-time cash flow data collected by Plaid.
In October, Plaid introduced a credit risk score that it said provides lenders with an up-to-date view of borrower risk by using real-time cash flow data and unique account connection insights from the Plaid Network.
Source: https://www.pymnts.com/