Capital One Reports 30% Developer Productivity Gains from Serverless-First Strategy

Quick Facts

• Capital One saves roughly 30% of developer time previously spent on infrastructure management
• The bank runs more than a third of its 2,000+ AWS applications on serverless infrastructure
• Capital One operates thousands of AWS accounts and tens of thousands of Lambda functions at enterprise scale

Capital One Financial Corp. reports significant productivity gains from its five-year transformation to a serverless-first operating model, with engineering teams saving roughly 30% of time previously spent managing infrastructure.

The financial services company now runs more than 2,000 applications on AWS, with over a third leveraging serverless infrastructure. Capital One operates thousands of AWS accounts and tens of thousands of Lambda functions, making it one of the largest AWS users alongside Netflix.

‘All the things that make the cloud great are enhanced by going serverless, which is a win-win for us and our customers,’ said George Mao, Senior Distinguished Engineer at Capital One.

Business Impact and Cost Savings

Teams adopting serverless services show significantly lower operational costs, according to Brian McNamara, distinguished engineer in Capital One’s Cloud Operations and Resiliency Engineering group. Developers no longer patch servers or rebuild operating system images.

‘To see our engineers focus more on delivering value to our customers than managing infrastructure’ drives the strategic shift, McNamara said.

The company established a Serverless Center of Excellence to define enterprise standards and best practices. Capital One uses a multi-account strategy where each team operates its own AWS account to limit blast radius and enable scaling with fewer constraints.

Market Context

The serverless computing market is expected to grow from $21.9 billion in 2024 to $44.7 billion by 2029 at a 15.3% compound annual growth rate, according to MarketsandMarkets Research.

Capital One follows a ‘serverless-first, not serverless-only’ approach, recognizing that very large workloads often run more economically on provisioned servers. The company closed its last data center in 2020, becoming one of only two Fortune 150 companies to go all-in on public cloud.

Capital One uses AWS Serverless Application Model (SAM) and canary deployments for Lambda updates. Using AWS Step Functions Distributed Map, the company processes thousands of workloads in parallel, significantly reducing check processing time.

Read more: Inside Capital One’s shift to a ‘serverless-first’ operating model