Terren Peterson, Capital One’s vice-president of engineering, knows the ins and outs of IT after 24 years at the bank. His perspective is simple: everything in IT eventually becomes commoditised.
“Back in the day, we spent countless hours dealing with IT infrastructure,” he reflects. “Now, everything’s in the cloud, and that infrastructure hassle? It’s mostly gone.” But while the nuts and bolts of IT have simplified, the challenges of managing data have only become trickier.
“Things like data management and problem-solving still need focus,” he stresses. “You can commoditise software and infrastructure, but data? That’s a different beast.” Peterson believes the trend will eventually mean fewer and fewer data management tasks. “Over time, those complexities will decrease, taking things off our plate.”
For Peterson, building a strong enterprise data architecture is crucial. It starts with fostering a data culture within an organization. Yet, many companies struggle with this. A recent survey showed only 35% of employees felt their workplace had a robust data culture. More than one in five reported inconsistent support for data initiatives.
“A data culture reflects discipline,” Peterson explains. “Are you making time for data management? Do you ensure data quality? Have you standardised your data?”
In a diverse company like Capital One, having a common language around data is vital. Peterson, as a data engineer, sees data platforms as key to this standardisation. “Platforms form the backbone of data engineering,” he points out. “Instead of everyone creating their own data pipelines, a data platform offers a shared solution.”
Peterson expresses excitement about the potential of AI and machine learning, noting, “That’s where we should focus our creativity—not on managing data.”
He champions the idea of generalising solutions. “When you can simplify the problem, you can create a platform everyone can use.” But there’s a challenge: getting team members to embrace this platform. “You need to channel their creative energy to the right areas,” he adds.
There are countless ways to set up a data pipeline, but Peterson advocates for standardisation. “Let’s not invent a hundred ways to do it.” Instead, he encourages creativity to find new data sources and explore innovative uses for data.
He likens the data pipeline concept to an enterprise service bus, which facilitates communication across applications. “You only need one service bus; the same goes for data platforms,” he says, aiming to streamline data standardisation so teams can unleash their creativity.
Peterson’s experience at Capital One exemplifies this strategy. He didn’t reinvent the wheel with a new data lake; he tapped into existing resources to extract unique insights. “Creative problem-solving is about leveraging what you already have,” he says, citing applications like improved fraud models or tools that help prospective car buyers.
Starting with a data platform requires a robust foundation. Many companies already have solid data practices. “Build on what you’ve got; it gives you a starting point,” he advises.
Using a tree analogy, he states, “The best time to plant a tree was 20 years ago. If you didn’t, today is the next best time. There’s no quick fix; you need that platform first.”
A recent example from Capital One is the Auto Navigator, launched in 2023. This tool connects dealers with buyers, helping streamline the car buying process by allowing customers to tailor payment details. It leverages comprehensive data across the business, requiring a strong data foundation and platform to support advanced analytics.
Ultimately, companies with a cloud-native IT strategy can swiftly implement a cloud-based data platform without the burden of configuring on-premise infrastructure.