Published in


From Raw Data to Actionable Insights: Introducing Prism Data and the CashScore

Petal was formed to fix deep-seated problems in the U.S. credit system. More than 50 million people lack a traditional credit score, and tens of millions of others have thin or stale credit information. These consumers are largely shut out of the financial system, unable to finance a car, purchase a home, or sometimes even find an apartment or a job.

When we started Petal in 2016, new sources of financial information like bank account transaction data — which is highly detailed and rich in financial insight — were quickly becoming accessible digitally. We believed these data sources could provide a useful substitute for traditional credit history. Further, emerging techniques in machine-learning (ML) and artificial intelligence (AI) would enable us to quickly and accurately analyze these millions of new data points, and with a more complete analysis, we could help millions of people gain access to safe and affordable credit. This belief kickstarted the work that would consume us for the next half-decade.

But just as we set out on our mission to democratize access to credit, we hit a roadblock: we first had to drastically improve the quality of the data.

Financial transaction data, which we refer to as “banking history,” is a disorderly and complicated mess. It’s inconsistent, mislabeled, and miscategorized, ultimately obscuring the financial realities the data can reveal. This messy data is nearly impossible to use for sophisticated analyses without significant work.

So we spent years building the software and models necessary to clean, categorize, and process banking history for complex use cases. We learned how to discern a consumer’s real financial story from the raw information — transforming a chronological list of transactions into useful and actionable insights, like the stability of their income, savings from month-to-month, and recurring financial obligations.

Only then were we able to build models, using ML and AI, to make predictions about credit risk. This resembled traditional cash flow-based underwriting, but instead of manual, paper-based and time-intensive analyses that might take weeks, we could make assessments in a fraction of a second. We honed our approach over several years: testing, learning, and improving our software and models.

This process was slow and expensive — in credit, learning means losing money — but the hard work and investment eventually paid off.

Today, Petal uses this technology to turn banking history into predictive scores for thousands of underserved consumers every month, facilitating access to hundreds of millions of dollars of credit. Automated cash flow underwriting has become a powerful and reliable substitute to traditional credit underwriting, creating an alternative credit score when traditional scores don’t exist, and improving the performance of traditional scores when they do. We’ve also developed proprietary software that uses banking history for income verification, customer identification (KYC), and fraud detection, among other things.

These are hard problems to solve, and we’ve come to appreciate they’re not unique to Petal.

As open banking accelerates, the use of comprehensive, real-time financial data has the potential to change the way consumer finance works — not just in credit, but in payments, banking, insurance, real estate, financial advising, and more.

That’s why we’re so excited to introduce our first B2B enterprise: Prism Data.

Prism Data makes Petal’s next-generation transaction intelligence platform available to the broader market for the first time, via a simple, lightning-quick API. Prism Data takes raw data from financial providers and transforms it into useful information that those providers can rely on, giving them greater insight into credit risk, identity, financial status, and more. We believe financial providers need actionable insights — not raw data — to create bold new solutions.

Prism Data provides advantages in underwriting, income verification, personal financial management, financial advising, fraud mitigation, tenant screening, and more, enabling partners to serve more customers, build better products, and make smarter decisions. We’re launching with four products to match this diversity of use-cases: Categories, Income, Insights, and alternative credit scoring, via our new “CashScore, by Prism Data.”

The CashScore incorporates all of Petal’s learnings about consumer cash flow underwriting, transforming banking history provided by a client or consumer into an accurate assessment of consumer credit risk.

No traditional credit history is necessary, which allows our platform to provide scores for thin-file and credit invisible consumers left out of the mainstream system. In our experience, no-file customers underwritten using the CashScore demonstrate credit performance equivalent to those with prime credit scores; when there is a full credit file available, the CashScore provides a differentiated, incremental view of risk, improving loss rates by 30%. The CashScore utilizes thousands of data points to assess risk, and runs on information that is up-to-date and nearly real-time. Importantly, it’s provided in a manner that is transparent to the consumer and compliant with consumer finance laws.

This innovation has never been more important. Over the past year, we’ve seen COVID-19 confuse traditional credit scores and cause large financial institutions to slam the brakes on lending. The CashScore can bring clarity to these decisions and help restore access to credit.

Prism Data is the next great step in furthering Petal’s original mission, now by empowering other organizations to serve more customers, build better products, and make smarter decisions. Just like light passing through a prism, raw, unfiltered information is organized into a clear spectrum of sophisticated financial insights. We’re excited to help our partners solve big problems and power the next generation of financial innovation.

If you’re interested in working with financial transaction data, we can help. We’re kicking off our limited beta with a few innovative partners eager for the future of open banking and transaction data intelligence. If you’re excited to learn more, or want to become one of our selected beta partners, let us know.




Thoughts and updates from Petal

Recommended from Medium

The Limits to Machine Learning: The Human Element in the Fake News Wars

Threat of AI on Human Jobs

Can We Escape an AI Arms Race?

Hard Lessons: What the World of Health AI Can Learn from Aviation

How AI Is Paving the Way for Autonomous Transportation?

Top 5 skills you must learn to tackle jobs in 2020.

Robotic Process Automation , Hyper Automation, Automation Hub , Hyper Automation Workforce

A Better Turing Test: How to measure the strength of AI

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Jason Gross

Jason Gross

Co-Founder & CEO, Petal.

More from Medium

Ridgeline In Review: “Find Those Who Believe What You Believe”

What gives startups the right to win?

Investing in Synnada

When ERPs Take a Back Seat