Uploaded April 2026 | Updated September 2026, 2 days ago
Viktor and Kostis dive into their experiences at KubeCon 2025, sharing insights on the overwhelming popularity of ArgoCon, which had to relocate to a larger venue mid-conference due to packed rooms. They discuss standout presentations including Kostis's talk on GitOps and secrets management, and highlight a compelling session from Adobe and IBM on using AI agents to automatically fix failed deployments and open pull requests. The conversation shifts to the evolving landscape of open source contributions in the AI era, addressing the challenge of AI-generated pull requests flooding projects and the responsibility maintainers have to guide contributors effectively.
The hosts tackle audience questions ranging from career development in DevOps and MLOps to technical decisions around infrastructure management. They debate the merits of multi-repo versus monorepo approaches, discuss Crossplane composition strategies using tools like Pythonic and KCL, and explore the balance between Terraform and Kubernetes-native solutions for infrastructure management. Viktor emphasizes his vision for the future of platform engineering, advocating for specialized agents integrated directly into developer tools like Claude and Cursor, while both hosts stress the importance of mastering fundamentals like Linux administration before pursuing specialized roles. Throughout the discussion, they reference key tools including Argo CD, Argo Rollouts, Crossplane, Chainsaw for testing, and various GitOps solutions.
▬▬▬▬▬▬ ⏱ Timecodes ⏱ ▬▬▬▬▬▬
00:00 Intro (skip to first question)
05:09 What are you guys working on these days?
06:36 AWS DevOps agent - have you tried it?
08:18 DevOps to MLOps to platform engineering focus?
12:23 AI helping deliver faster - preventing security issues?
17:00 Crossplane composition organization - monorepo vs multirepo?
20:41 Managing fleet of Argo CD clusters across clouds?
23:45 AI-generated pull requests in open source projects
26:58 Crossplane composition functions - why choose Patroni?
27:58 Management cluster with Argo and Crossplane syncing strategy?
40:13 Full-stack developer diving into DevOps and infrastructure?
43:47 Open source community becoming harder with AI?
49:40 Using Kubernetes to manage other Kubernetes clusters?
56:11 Terraform for bootstrapping vs Kubernetes-native approach?
58:11 How to learn AI ML Ops?
1:00:44 What creates and manages the management cluster?
1:03:35 Best platform for lean team without ops experience?
1:04:53 Where to find remote DevOps jobs?
1:05:56 What's next in platform engineering?
1:08:17 Argo CD agent project replacing existing management approaches?
Viktor and Kostis dive into their experiences at KubeCon 2025, sharing insights on the overwhelming popularity of ArgoCon, which had to relocate to a larger venue mid-conference due to packed rooms. They discuss standout presentations including Kostis's talk on GitOps and secrets management, and highlight a compelling session from Adobe and IBM on using AI agents to automatically fix failed deployments and open pull requests. The conversation shifts to the evolving landscape of open source contributions in the AI era, addressing the challenge of AI-generated pull requests flooding projects and the responsibility maintainers have to guide contributors effectively.
The hosts tackle audience questions ranging from career development in DevOps and MLOps to technical decisions around infrastructure management. They debate the merits of multi-repo versus monorepo approaches, discuss Crossplane composition strategies using tools like Pythonic and KCL, and explore the balance between Terraform and Kubernetes-native solutions for infrastructure management. Viktor emphasizes his vision for the future of platform engineering, advocating for specialized agents integrated directly into developer tools like Claude and Cursor, while both hosts stress the importance of mastering fundamentals like Linux administration before pursuing specialized roles. Throughout the discussion, they reference key tools including Argo CD, Argo Rollouts, Crossplane, Chainsaw for testing, and various GitOps solutions.
▬▬▬▬▬▬ ⏱ Timecodes ⏱ ▬▬▬▬▬▬
00:00 Intro (skip to first question)
05:09 What are you guys working on these days?
06:36 AWS DevOps agent - have you tried it?
08:18 DevOps to MLOps to platform engineering focus?
12:23 AI helping deliver faster - preventing security issues?
17:00 Crossplane composition organization - monorepo vs multirepo?
20:41 Managing fleet of Argo CD clusters across clouds?
23:45 AI-generated pull requests in open source projects
26:58 Crossplane composition functions - why choose Patroni?
27:58 Management cluster with Argo and Crossplane syncing strategy?
40:13 Full-stack developer diving into DevOps and infrastructure?
43:47 Open source community becoming harder with AI?
49:40 Using Kubernetes to manage other Kubernetes clusters?
56:11 Terraform for bootstrapping vs Kubernetes-native approach?
58:11 How to learn AI ML Ops?
1:00:44 What creates and manages the management cluster?
1:03:35 Best platform for lean team without ops experience?
1:04:53 Where to find remote DevOps jobs?
1:05:56 What's next in platform engineering?
1:08:17 Argo CD agent project replacing existing management approaches?
