azCloudOps

Between November 9 and December 4, the Cloud Native Computing Foundation, Microsoft and AWS each hold their biggest event of the year. What lands in those four weeks will shape most cloud and AI roadmaps for 2027. Here is what operations teams should watch at each, and how to follow along without attending.

Event	Dates (2026)	Where	Remote access
KubeCon + CloudNativeCon North America	November 9 to 12	Salt Lake City	No virtual component; session recordings posted to the CNCF YouTube channel within two weeks
Microsoft Ignite	November 17 to 20 (Security Pre-Day November 16)	San Francisco	Online attendance available
AWS re:Invent	November 30 to December 4	Las Vegas	Free livestream of keynotes and more, with registration

KubeCon NA: AI workloads become a platform problem

What changed. The CNCF’s own guide to the week shows where the cloud native community’s attention is. The Monday co-located program includes Agentics Day: MCP + Agents, Cloud Native AI + Inference Day, Observability Day, Platform Engineering Day and Open Source SecurityCon, alongside project events such as ArgoCon, BackstageCon and CiliumCon. The CNCF lists AI-aware networking, GPU and specialized device management, AI and ML workload orchestration, and agent observability among the session topics. The co-located events need an All-Access Pass.

Where it is heading (analysis). Inference is moving from dedicated research clusters onto the shared platforms that platform teams already run. That shifts GPU scheduling, multi-tenant isolation and cost allocation from the ML team’s problem to the platform team’s problem. Expect the Kubernetes conversation in 2027 to be about running inference as a well-behaved tenant, not about whether to run it at all.

Watch for:

• Patterns for GPU sharing and scheduling on AKS, EKS and GKE, and which ones are upstream rather than provider-specific.
• How agents and MCP servers are being deployed, secured and observed on Kubernetes.
• Agent observability work, and whether it converges on OpenTelemetry conventions.

Microsoft Ignite: governing agents at enterprise scale

What changed. Microsoft’s security guide to Ignite organizes its track around four themes: Defend with AI, Secure AI, Get Ready for AI, and Secure Data. It names an inside look at Microsoft Security as customer zero for Agent 365 and a hands-on lab in agentic security, and its Secure AI track covers securing AI runtime with Microsoft Defender and data protection and governance for agents with Microsoft Purview. A Security Pre-Day runs on November 16.

Where it is heading (analysis). The agenda signals that Microsoft treats agents as a governed workload, with identity, data protection and runtime security applied to them the way they are applied to users and devices. For organizations on Microsoft 365 and Azure, the operating question for 2027 is less “can we build agents” and more “who registers, approves, monitors and retires them.”

Watch for:

• Agent 365 announcements, and in particular any change in availability, licensing or how it relates to existing identity and compliance tooling.
• How Defender and Purview coverage extends to agents that act across services, and what telemetry operations teams will get.
• Azure infrastructure and cost announcements for AI workloads, which tend to arrive alongside the headline Copilot and agent news.

AWS re:Invent: launches, migration, and the cost of AI

What changed. AWS describes re:Invent 2026 as “cloud and AI’s biggest week,” with the headline “Build what’s next.” The event page promises over 2,200 sessions across AI, security, migration and developer productivity, framed around an architecture decision, a migration tradeoff, token spend or security risk. AWS’s cloud financial management team has published a guide to more than 30 cost sessions, including FinOps for AI: Cost Allocation Strategies for Amazon Bedrock, plus a December 1 FinOps Foundation program with the Tokenomics Foundation and the FOCUS team.

Where it is heading (analysis). When a cloud provider puts token spend in the description of its flagship conference, AI cost has become a mainstream operating concern rather than a FinOps specialty. Expect more native tooling for attributing and capping AI spend, and pressure on every provider to report AI usage in standard billing formats. Our deep dive on FOCUS 1.5 and token-level cost allocation covers what that format is set to look like.

Watch for:

• New AI and agent services, and the governance controls that ship with them on day one versus later.
• Cost management announcements for AI and token spend, and whether they align with the FinOps Foundation’s FOCUS specification.
• Migration tooling and programs, which matter for anyone consolidating estates in 2027.

The thread running through all three

Read side by side, the three agendas point at the same operating gap: AI workloads and agents are arriving faster than the platform, security and cost controls around them. KubeCon approaches it from the platform, Ignite from identity and data protection, re:Invent from services and spend. Organizations that run more than one cloud will need one answer that works across all three, which is exactly the conversation this community exists to have.

What to do before the season starts

• Agree as a team which two or three questions you need answered by these events, so you watch for decisions, not headlines.
• Assign one person per event to watch the keynote and the security and cost sessions, and to write five lines of takeaways.
• Hold anything you were about to commit to (a reservation, a tool purchase, a platform redesign) until the relevant event has passed, if the decision could plausibly change.
• Book a one-hour review in the second week of December to turn the takeaways into 2027 roadmap items.

Sources

• CNCF: KubeCon + CloudNativeCon North America 2026, your week in Salt Lake City (September 11, 2026)
• Microsoft Security Blog: Secure what’s next, your guide to Microsoft Security at Microsoft Ignite 2026 (September 30, 2026)
• AWS re:Invent 2026 event page
• AWS Cloud Financial Management Blog: Know Before You Go, your re:Invent 2026 guide to managing cloud and AI spend (October 16, 2026)

Last verified: November 5, 2026.