What does 20 years at CreditXpert look like?
Just ask Peter!

This week, fan-favorite Peter Fitton, celebrated 20 years of employment with CreditXpert! He joined as a data scientist, then was promoted to manage other data scientists, then Product Manager, onto Analytics & Research Manager, and is now the Director of Business Analytics. With that kind of tenure and success, comes a wealth of knowledge about the industry and of the organization. We asked him a few questions to try to get the secret to success and how to stay motivated in the workplace.

20 years is a long time, what has kept you here for so long?

“I really believe in the mission. To my mind, credit scoring is fundamentally unfair to consumers in the sense that they are evaluated against a scorecard but the scorecard is secret. All the generic credit advice they read only helps keep their credit in good shape if it is already good. It doesn’t tell them how to recover from a life event that negatively impacted their credit score, which most people will experience some time in their lives. Our tools give consumers a specific plan to follow to help their credit scores recover, and mortgage companies pay us so that they can provide that plan to a consumer for free. That’s doing something amazing for consumers.

I like my coworkers! We do a good job of hiring nice people who are good at their jobs and live out our company values. The pay and especially benefits are very good. I feel like my family is well taken care of.”

How has data science changed in your 20 years at CreditXpert?

“When I first started at CreditXpert, few people were using the term “data science”. We almost did… my title was Senior Scientist, with the “data” part implied rather than an explicit part of the title. Back then we had what were high-end PCs, but it still took us a day or two to train a single-hidden-layer neural network on a dataset consisting of a few tens of thousands of records (something that would now take a few minutes). Our data science tech stack at that time consisted of SAS, a homebrew neural network tool, and Excel, all running on-prem. Machine learning models could only be deployed as custom code.

As storage and computing power became cheaper, and especially after AWS and other cloud computing services took off in the late 2000s and the use of GPUs for matrix calculations became more common, both the data available and the means to analyze and model it scaled way up. As a result, the demand for data scientists across many industries went way up in the early 2010s, and more people entered the field. The mid-to-late 2010s was the era of big data and machine learning, and everything moved to the cloud. Our tech stack now consists of cloud-based technologies like Snowflake, Fivetran, dbt Cloud, Looker, and Dataiku. The ability to push down nearly all computations into our Snowflake database gives us an enormous amount of power, but only when we need it. The field continues to advance rapidly, be applied to more business problems, and attract more people.”

Looking to the next 20 years, how do you see see data science being able to help CreditXpert fulfill on its mission of “making homeownership more accessible and affordable for all?”

“Over the next few years, we expect the mortgage industry to adopt a new set of credit scoring models, and we will use data science to support that transition. Besides that, there is a lot of opportunity to use data science to improve the user experience of our products. The more we apply intelligent automation within our software, the easier it becomes for loan originators to help more applicants, for applicants to be successful in improving their credit, and for decision-makers to measure the value our products deliver.”

Thank you Peter, for your generous 20 years at the CreditXpert — here’s to 20 more!