Tech Tuesdays: AI Agents, Modernising Code and the Future of DevOps скачать в хорошем качестве

Tech Tuesdays: AI Agents, Modernising Code and the Future of DevOps

Welcome to another Tech Tuesdays episode of Vocal Technologist. Today we are exploring how artificial intelligence is moving beyond simple code completion and becoming a core part of our platforms. Agentic platform engineering is emerging as the next evolution, allowing platforms to handle complex tasks with bounded autonomy while humans provide oversight. The industry is taking this seriously, with former GitHub CEO Thomas Dohmke raising sixty million dollars to rebuild the software development lifecycle specifically for AI agents. Our everyday development tools are rapidly evolving to support this shift. Visual Studio Code has transformed into a multi agent command centre that can orchestrate several different models at once. Docker has also stepped up, launching shell sandboxes that let developers run AI assistants in isolated environments with secure credential management. Furthermore, teams using Azure Boards can now integrate custom agents tailored to their workflows. However, while new technology accelerates, developers are still wrestling with legacy systems and technical debt. Java developers are feeling the weight of vulnerability management and licensing costs, prompting many to migrate to OpenJDK. Fortunately, tools like AWS Transform custom are using AI to automate these massive Java modernisations. The Go community is also tackling code health by rewriting their tooling to automatically modernise syntax, while Prometheus is upgrading its storage systems to natively support composite types. On the infrastructure side, operations teams have exciting new capabilities. AWS has introduced nested virtualisation for specific instances and open sourced a controller to improve Kubernetes high availability. We are also seeing smarter infrastructure as code, with Pulumi enabling secure database authentication and integrating Terraform state outputs. For highly secure setups, Google Distributed Cloud has brought public cloud networking features to air gapped environments. Finally, we discuss how teams measure all this progress. A surprising number of platform teams fail to measure their success, risking their funding if they cannot demonstrate business value using metrics like DORA. The only productivity metrics that truly matter are how often a team ships and how rarely they cause production outages. Fast teams maintain a bias toward action by making small deployments safe and reversible. https://vocaltechnologist.cyou