Cloud and AI now change faster than most teams can read about them. azCloudOps is a vendor-neutral place to read, discuss and decide what those changes mean for the people who have to operate them, across Azure, AWS, Google Cloud and the AI services running on all three.
Vendor blogs announce features. Analyst reports sit behind paywalls. Practitioner discussion is scattered across LinkedIn, Reddit, Discord and conference hallways. Each platform, operations, security and FinOps team ends up rebuilding the same understanding on its own. This site is an attempt to do that work once, in the open, with the people doing the work.
Four questions, every time
Every article on azCloudOps answers the same four questions, in the same order:
- What is it? Plain language and an everyday comparison before any jargon.
- What changed? The verified update, its status (preview, generally available, retiring) and a link to the vendor’s own documentation, specification or announcement.
- Where is it heading? The next six to eighteen months. This part is analysis, and it is labeled as analysis.
- What should I do? Next steps by role, written so you can hand them to someone this quarter.
The A to Z: six pillars
The name is a promise of range. Coverage is organized into six pillars, and every article sits in exactly one of them:
| Pillar | What it covers |
|---|---|
| Architecture and Landing Zones | Account and subscription design, networking, infrastructure as code, migration |
| Operations and Reliability | Observability, SRE practice, incident response, AIOps |
| Security, Identity and Governance | Identity, policy as code, compliance evidence, controls for AI agents |
| FinOps and Cost | Allocation, commitments, chargeback, AI and token spend |
| Platform Engineering and Kubernetes | Internal platforms, GitOps, AKS, EKS, GKE, inference on shared clusters |
| AI and Agent Operations | MLOps, LLMOps, evaluation, model and agent lifecycle |
Alongside the articles, the Journal holds my own notes: experiments, conference takeaways and lessons learned.
What to expect, and when
- Every Thursday: one new article, usually with a diagram and next steps by role.
- First week of each month, starting in December: Roadmap Radar, a summary of what moved across the cloud and AI roadmaps last month and what to do about it.
- Around major events: an Event Brief before and after, starting with KubeCon, Microsoft Ignite and AWS re:Invent this November.
- From the community: Field Notes written by practitioners, through the proposal process below.
Two pieces launch alongside this one: a guide to the three conferences in the next four weeks, and FinOps for AI and what the FOCUS specification gives you now. They join a deep dive on FOCUS 1.5 and token-level cost allocation.
This is a community, not just a blog
Articles are the starting point. The useful part is what happens after: the operator who has already hit the limit, the architect who disagrees with the analysis, the FinOps lead who has a better allocation model.
- Comments are open on every article, for questions, disagreement and your own experience.
- Contribute to propose an article, suggest a topic or report an error. If you run systems on any of the three clouds, you have something worth writing about.
Please read the Code of Conduct before you take part.
The standards
- Every product claim links to a primary source.
- Preview features are marked as preview. Forward-looking statements are marked as analysis.
- Articles carry a last-verified date and are corrected in place, with the change noted.
- No sponsored content. Contributors disclose their employer when they write about its products.
Start here
Introduce yourself in the comments below: your role, the clouds you run, and the one operating problem you would most like this community to help solve. Then subscribe to get each new article by email.
Manoj Annavajjala, founding editor
Leave a Reply