Cognition Partners With AWS to Deploy Devin for Enterprises

Cognition and Amazon Web Services have signed a multi-year Strategic Collaboration Agreement to make Devin, an autonomous software engineering agent, available via the AWS Marketplace. The partnership gives enterprise customers a standardized channel to deploy AI software engineers directly into production environments.

"For most of software's history, engineers have been stifled by the need to work on legacy systems. That changes with Devin," said Art Levy, VP of Global Partnerships at Cognition.

Joint customers can purchase Devin through the marketplace and utilize the Agent Toolkit for AWS directly within the platform. The agreement pairs Cognition's flagship agentic AI software with Amazon's cloud infrastructure to scale enterprise application development and modernization workloads.

Security Architecture and Tenant Isolation

Enterprises deploy Devin within a single-tenant environment on AWS, allowing organizations to select their preferred operating region. Data and internal organizational knowledge remain secured inside a dedicated Virtual Private Cloud, while each individual session operates within its own isolated virtual machine.

"Our customers want to modernize faster and put AI to work in production, and Devin running on AWS gives them a proven way to do it," said Anasuya Strasner, Director of North America ISVs at AWS.

This infrastructure design addresses corporate data governance requirements by keeping sensitive codebases out of shared or multi-tenant pools during automated debugging and development tasks.

Early Case Studies in Legacy Modernization

Automated coding agents have increasingly targeted legacy codebases, which consume substantial maintenance resources across banking, logistics, and manufacturing sectors. Enterprise deployments detailed in the partnership announcement show significant efficiency gains in mainframe migration projects.

"Devin has become an important part of how our engineering teams work at Echo, from DeepWiki-powered code insights that help us understand our codebases, to closing gaps in legacy systems," said Christina Garcia, SVP of Engineering at Echo Global Logistics.

Mercedes-Benz utilized Devin to analyze more than 200,000 lines of COBOL code, compressing an estimated eight-month modernization timeline into eight days. Separately, an unnamed global automaker used the agent to migrate a 25,000-line COBOL workflow from an on-premises mainframe to AWS Lambda, achieving an estimated 73% reduction in operating costs.

What Is at Stake for Enterprise Software Development

The multi-year alliance positions cloud providers and AI startups to capture growing corporate IT spending directed toward automated software maintenance. By embedding autonomous agents directly into managed cloud environments, technology vendors aim to lower the deployment friction associated with legacy code conversion and large-scale codebase analysis.