Azure Kubernetes Service (AKS)
How Azure Kubernetes Service works: AKS Automatic and Standard, pricing tiers, Azure CNI networking, Workload ID, Gateway API ingress and upgrade channels.
Cloud · AI · Generative AI
Complex tech, broken into fun, digestible bites. Practical guides to AWS, Google Cloud and Azure, and plain-English tours of machine learning and generative AI.
Fresh from the blog
Learning paths
Three series that build on each other: start with the cloud, meet the machines that learn, then watch them create.
Cloud computing
Foundations of the cloud and hands-on tours of AWS, Google Cloud and Azure — from your first VM to multi-cloud strategy.
AI & machine learning
How machines learn: the history of AI, supervised and unsupervised learning, model training and the data behind it all.
Generative AI
Transformers, GANs, VAEs and GPT — what generative AI is, how it works, and how the cloud providers put it to work.
Quick quiz
Three quizzes, 150 questions, instant explanations. Your best scores stay on your device.
The 'shared responsibility model' in cloud computing primarily defines the security obligations between:
The shared responsibility model outlines which security tasks are handled by the cloud provider (e.g., security OF the cloud infrastructure) and which are handled by the customer (e.g., security IN the cloud, like data configuration and access management).
Keep exploring
How k-nearest neighbors classifies by looking up similar examples: distance metrics, choosing k, scaling, the curse of dimensionality and fast neighbor search.
How Amazon ECS and Amazon EKS run containers on AWS, what Fargate, Managed Instances and Auto Mode change, and how to choose between them.
How CycleGAN learns to turn horses into zebras and photos into paintings without paired data: cycle consistency, the losses, the architecture and PyTorch code.
How support vector machines draw the widest possible boundary between classes: margins, support vectors, kernels, C and gamma, with working scikit-learn code.
What a container actually is, how it differs from a VM, and how to build, run and ship your first image with Docker before moving it to the cloud.
How StyleGAN reaches 1024 by 1024 faces: the mapping network, per-layer styles, noise inputs, the truncation trick, and what StyleGAN2, ADA and StyleGAN3 changed.