Tool Open source
Kubernetes (K8s) is an open-source container orchestration system hosted by the Cloud Native Computing Foundation. It manages containerized applications across multiple hosts, providing mechanisms for deploying, maintaining, scaling, and scheduling workloads. Applications can be configured with YAML and deployed across hybrid-cloud environments; the platform is also used for production operations involving storage, databases, networking, and security hardening.
20 uses taken from transcripts — each links to the moment in the video.
Use YAML configuration to deploy containerized applications to a hybrid cloud.
Reference model for a control plane that manages distributed workloads, adapted here to agent identity, policy, observability, lifecycle, and execution.
A container orchestration platform discussed for production deployments, DevOps experience, virtual clusters, database operations, storage, and security hardening.
Deploys the self-hosted project-management platform.
Used as an example platform for running multiple copies of a service to improve resilience, while potentially increasing system complexity.
Support deployment of containerized AI agents and models across environments including hybrid cloud.
Hosts the production deployment, which should run behind an authentication proxy.
Infrastructure platform used while migrating the WordPress blog.
Runs the language-model workload, GPU nodes, pods, services, ingress, deployments, and autoscaling configuration.
Hosts the remote coding container in a cluster and provides deployment health signals that can be checked before production release.
Runs the product and data-processing pipeline across cloud, on-premises, and other environments.
Runs underneath Slurm to provide infrastructure resilience, node health management, observability, automatic healing, load balancing, autoscaling, and shared GPU capacity for training and inference.
Cluster runtime for the self-hosted deployment
Spins up a pod per agent session that is destroyed when the run finishes; cloud scale to thousands of concurrent agents, with cron jobs and HTTP event triggers for full automation.
Runs the microservices as three independent clusters, one per cloud provider (GCP, Azure, AWS)
Production deployment platform for agents in a namespace
Google's open-source container orchestration system, code-named Project Seven by Joe Beda, Brendan Burns and Craig McLuckie. It lets you declare desired state (CPU, memory, replicas) and a control loop keeps the cluster matching it; donated to the CNCF in 2015 and now the industry-standard way to package and run software.
Compute target customers may require for agent sandboxes
Cited as infrastructure whose pod restarts can force a job retry, motivating shared payment state
Platform on which vLLM instances are scaled and llm-d routes inference requests.
20 in the library.