How I learned to stop worrying and love chaos (Engineering)
In this episode of the AWS Developers Podcast, recorded live at the AWS Summit in London, we dive into the world of chaos engineering with guest Simon Hanmer, Principal Consultant at GlobalLogic and AWS Community Builder. Together with Tiffany and Sébastien, we unpack what chaos engineering is, why it matters for resilience in modern cloud architectures, and how AWS customers are adopting these practices today.
Simon explains how chaos engineering isn't about breaking things for fun, but about building confidence in how systems behave under stress—just like astronauts or firefighters train for the worst-case scenarios. We discuss AWS Fault Injection Service (FIS), best practices for injecting controlled failures, and how to safely test your assumptions before disaster strikes. Simon shares practical insights from his work with enterprise customers, the evolution of resilience testing from data centers to the cloud, and what’s next for chaos engineering, including integrating into CI/CD pipelines and shifting testing left.
If you're curious about how to make your cloud architecture truly resilient—or how Netflix and Amazon do it in production—this is the episode for you.
--------
38:46
Navigating Machine-to-Machine Security
In this episode, Seb and Abram Douglas dive deep into OAuth 2.0 and the challenges of machine-to-machine (M2M) authentication. They unpack the security trade-offs between API keys and the client credentials grant flow, explaining how Amazon Cognito can generate time-bound access tokens and use Lambda triggers for token customization. The conversation highlights token claims, secure verification methods, and how API Gateway integrates with Cognito for simplified authorization. Seb and Abram also explore fine-grained access control using Amazon Verified Permissions and outline best practices like securing secrets with AWS Secrets Manager, rotating client credentials, and enabling AWS WAF. Finally, they look ahead to the role of AI agents in secure M2M communication, stressing the importance of user consent, identity propagation, and robust token management in future architectures.
--------
34:15
Adobe: 10 years of cloud, how to not clutter your AWS accounts
In this episode, we celebrate a decade of cloud innovation at Adobe with Johannes Gehrs, Site Reliability Engineer, who shares how his team keeps AWS environments clean, efficient, and developer-friendly. We dive into practical strategies to reduce account sprawl, enforce good governance, and standardize everything on infrastructure as code. From using service control policies to building an internal developer platform, Johannes brings concrete tips that scale across large organizations. Whether you're wrestling with cloud governance or looking to optimize your AWS spend, this episode delivers insights you can apply right away.
--------
32:21
Cancer Research UK goes Serverless
In this episode of the AWS Developers Podcast, we explore the first major step in Cancer Research UK’s broader move to serverless: the migration of their payment systems. While this application was the first to make the shift, it paved the way for many other products and services that have since followed suit—marking a significant milestone in their serverless journey. Our guests, Ellie Wintram and Swetha Podduturi, walk us through the motivations behind the migration, the architecture decisions that shaped it, and the key AWS services that made it possible. They share how technologies like AWS Lambda and DynamoDB enabled cost savings of up to 94%, why TypeScript was the right fit for their microservices, and how observability and load testing were central to building trust in the new system. Whether you're planning a migration or curious about serverless best practices in production, this episode is packed with technical insights and practical lessons from the field.
--------
30:32
Ryanair: a journey towards responsible AI
In this episode of the AWS Developer Podcast, we dive into Ryanair's journey towards responsible AI. As the airline aims to carry 300 million passengers in less than a decade, it is turning to generative AI to enhance productivity and scale operations. The conversation explores how Ryanair is identifying and implementing use cases, the role of its generative AI center of excellence, and the importance of embedding governance into AI from the very start. You'll hear how the airline is balancing agility with compliance through the creation of an AI compliance hub and a flexible, evolving framework to manage AI risks. The speakers share candid insights into lessons learned, the challenges of cross-functional collaboration, and why responsible AI is a continuous process—one that future teams will be grateful for.
UK