How Lockheed Martin Builds Software Faster Powered by Devsecops Dop323

Title

AWS re:Invent 2023 - How Lockheed Martin builds software faster, powered by DevSecOps (DOP323)

Summary

  • Lockheed Martin, the world's largest defense contractor, has undergone a transformation journey to improve its software development process.
  • Alan Horn, director of software security at Lockheed Martin, discussed the company's transition from a slow, methodical approach to a more rapid iteration process while maintaining mission-critical and safety-critical standards.
  • The company faced challenges due to its size, geographic distribution, and the classified nature of its work, which required isolated, closed spaces for software development.
  • Lockheed Martin adopted a bottom-up approach to transformation, leveraging feedback from passionate individuals across the business to drive change.
  • The company moved to a common DevSecOps platform using GitLab and AWS, which allowed for scalable, automated builds and deployments, improving deployment frequency and reducing risk.
  • Key metrics for success included the DORA metrics (deployment frequency, change failure rate, mean time to repair) and the number of users on the platform.
  • Future visions include leveraging sovereign cloud regions for global operations and incorporating generative AI to enhance developer productivity and system design.

Insights

  • Lockheed Martin's transformation journey emphasizes the importance of a bottom-up approach, where feedback from the front lines is crucial for successful change.
  • The company's move to a common DevSecOps platform highlights the benefits of a unified approach to collaboration and asset management.
  • The adoption of GitLab and AWS has led to significant improvements in deployment frequency and scalability, demonstrating the effectiveness of cloud-native architectures.
  • Security is deeply integrated into the development process, with security experts actively participating in the DevSecOps lifecycle.
  • The discussion on generative AI suggests that organizations need a mature software engineering lifecycle foundation to fully benefit from AI advancements.
  • The transformation journey at Lockheed Martin serves as a case study for other large, geographically distributed organizations facing similar challenges in software development and security.