All posts
Case Studies

How Capital One Left Eight Data Centers for AWS, and What the 2019 Breach Taught Everyone

By

October 25, 2026

7 min read

Capital One was the first US bank to announce going all in on the cloud, closing eight data centers by 2020 and rebuilding 80% of about 2,000 applications as cloud native. What changed in delivery speed and reliability, and the security lesson from 2019 that changed EC2 for every customer.

Case Study
Financial Services
Cloud Migration
Security
Capital One

What they did

  • An eight-year program, completed in 2020, that exited eight data centers.

  • About 2,000 applications moved, and 80% were rebuilt as cloud native (microservices and APIs) rather than lifted and shifted.

  • The technology organization grew to about 11,000 people during the migration. This was a hiring and skills program as much as an infrastructure one.

Want a second opinion on this for your system?

Book a free 30-minute call. You'll leave knowing what we'd do first and roughly what it costs, whether or not you hire us.

What changed, by AWS's account

Measure

Before

After

Building a development environment

About three months

Minutes

Disaster recovery testing time

Baseline

70% less

Time to resolve critical incidents

Baseline

50% less

Transaction errors

Baseline

50% fewer

Release frequency

Monthly or quarterly

Many times a day

The 2019 breach

In 2019 an attacker exploited a misconfigured web application firewall to make requests from inside Capital One's environment to the EC2 instance metadata service, obtained the instance role's credentials and used them to read data from S3. That November AWS released IMDSv2, which requires a session token and blocks this class of server-side request forgery. Today IMDSv2 can be required across an organization by declarative policy.

Our read

  • Rebuilding most applications instead of rehosting them is why delivery speed changed. A rehosted application keeps its release process.

  • It took eight years and thousands of engineers. Smaller companies get the same outcome faster by modernizing the few applications that change most and rehosting or retiring the rest.

  • The breach was not a cloud failure; it was a configuration and permissions failure that the cloud made reachable. Least-privilege roles, IMDSv2 and data perimeters exist because of incidents like it. Build them in from the first account.

hoolco Monthly Review

Once a month: what changed in tech that month and what it means for you, from cloud and AI to security and the tools you build with. Every item links to its source. No sales pitch, unsubscribe in one click.

As of October 25, 2026. Provided as is, for general information, and accurate to the best of our knowledge on the date shown. Not legal, compliance, tax, financial or other professional advice. Prices, limits and vendor services change; check current vendor documentation before relying on anything here. Terms of use.