KubeCon 2025
The second KubeCon + CloudNativeCon in India. Two days in Hyderabad: Kubernetes at scale, cloud-native AI, and the people who make it worth showing up in person.
KubeCon + CloudNativeCon India was back in August 2025, this time in Hyderabad, and there was no chance I was missing it. Two days, the back half of the same trip that opened with OpenSSF Community Day, but at a completely different scale: the full cloud-native community under one roof, more than a dozen tracks running at once, and a keynote hall packed to the walls. I spent most of it doing the thing conferences are actually for.
The reason you actually go
The talks get recorded and go up on YouTube within weeks; the reason to fly to Hyderabad is everyone else in the building. The keynote hall was packed wall to wall, and the hallway track was better still. I finally met Mumshad Mannambeth, whose courses have quietly certified a huge share of the room, mine included, and spent the two days trading notes with maintainers, CNCF ambassadors, and user-group organisers from across Asia, the KSUG crew from Singapore among them. That is the part that does not get recorded, and the part I go for.
At some point someone called for a group photo, and the community actually assembled: engineers from every corner of the country and well beyond, squeezed into one hotel corridor.
What I actually went for
The schedule ran more than a dozen tracks in parallel across the two days, so nobody saw even a tenth of it. This is my slice, the sessions I picked out of the overlap because they map to what I actually work on. Three threads ran through all of it: Kubernetes as the operating system for AI, from LLM inference to GPU observability; platform engineering as a real discipline, with internal developer platforms, Backstage, Karpenter, KRO, and ephemeral environments on EKS; and security woven through the lot. Each links to its page on the schedule.
- Keynote: Kubernetes for GenAI: The New Operating System of the AI-First WorldJanakiram MSV, Janakiram and Associates
- Keynote: Pushing the Limits: Scaling ArgoCD's Repo-Server To Handle 30K ApplicationsAditya Sharma & Isha Gusai, Expedia Group
- Supercharge Backstage: Self-Service GitOps Staging Environments With VCluster and FluxCDFarhaan Shamsee & Rakesh Reddy, Siemens Technology
- Guard Your Network With Kyverno and EnvoySanskar Gurdasani, Accuknox & Swastik Gour, InfraCloud
- To Succeed in Platform Engineering, Build Infrastructure Knowledge GraphsSafeer C M, MoEngage
- Introducing Kubernetes Resource Orchestrator (KRO)Abdel Sghiouar & Mofi Rahman, Google Cloud
- IDP as a Product: Where Developer Happiness Meets Platform's ExcellenceNinad Desai & Ruturaj Kadikar, InfraCloud Technologies
- Scaling AI Like a Pro: PepsiCo's LLM Deployment Strategy on KubernetesPraseed Naduvath & Dhanashree Shetty, PepsiCo
- Building a Cloud Native IaC Platform for an EnterpriseFlorian Hopfensperger, Allianz Technology & Yury Tsarev, Upbound
- Rewriting the Rules: Why Kubernetes Is Moving To NFTablesDaman Arora & Yash Kumar Singh, Broadcom
- A Journey To Ephemeral Test Environments With EKS and ArgoCDTatiana Selezneva & Gazal Gafoor, REA Group
- Keynote: From Outage To Observability: Lessons From a Kubernetes MeltdownSaiyam Pathak, LoftLabs & Arnab Chatterjee, Nomura
- Keynote: Intuit's AI-Native Platform: Scaling to 8x Velocity for 8,000 DevelopersJimil Patel, Intuit
- Keynote: Kubernetes and Platform Engineering: A Strategic Partnership for Scalable InnovationBhavani Indukuri, Zscaler
- Handling Node Churn in Karpenter: Efficient Scaling for Large EKS ClustersShivani Mehrotra, Expedia Group & Chetan Saini, Antra
- Capture the Flag ExperienceControlPlane
- Cloud Native GenAI Using KServe and OPEAJohnu George & Gavrish Prabhu, Nutanix
- Cluster API To Streamline Kubernetes Clusters Lifecycle Management at NVIDIAMegha Singhal & Gulshan Khatri, NVIDIA
- Auto-instrumentation for GPU Performance Using eBPFAnnanay Agarwal, Grafana Labs
- Fine-Grained Authorization: The Missing Piece in Agentic AI SecurityShivay Lamba, Couchbase & Ashish Jha, Okta
- How Intuit Streamlined AI/ML Inference Workflows on K8sYashash H L & Sreekanth P R, Intuit
- Observability at Scale With Monitoring as Code: Grafana, Prometheus, and TempoVipin GopalaKrishnapillai & Saiabhinay Bommakanti, Amway
- Multi-Layered Guardrails for Cloud Native AI: Enforcing Compliance and Safety at ScaleVincent Caldeira & Anindita Sinha Banerjee, Red Hat
- Sandboxing Agentic AI With LSM-BPFRahul Jadhav, Accuknox
Collaboration, out in the open
The best sessions were the unofficial ones. Put a few engineers around one laptop and within minutes it turns into a working session: someone drives, everyone else debugs over their shoulder. A lot of what I took away from the two days happened around tables like this one, nowhere on the schedule.
Between talks the corridors did the real work. Conversations started with 'so what are you running?' and ended with a repo link, an introduction, or a friendly argument about etcd. I had more of those in two days than in months of remote work.
And some of it was simply catching up: people I had only ever known as GitHub handles and Slack avatars, finally attached to faces, usually somewhere near the buffet.
