Book a demo

Platform Engineering Beyond the Hype

Get valuable insights on Platform engineering, Kubernetes operations, cloud automation, and enterprise application platforms with the following videos.
YouTube Video UEwxWXdxaGkyb3JOMGxxVERhNUxHUG1BdmdUMVRlWWhncy5FQjcyRDZBMEMyNTY3OTdE Most platform teams reach for Crossplane when they need to manage data services across Kubernetes clusters — and most of them hit a wall. The moment you introduce dozens of databases, hundreds of tenants, and thousands of service bindings, the overhead becomes unsustainable and self-service for developers breaks down completely. Julian Fischer, CEO of anynines, explains exactly why Crossplane alone can't solve this problem — and how Klutch, the open source project at the core of a9s Hub, fills the gap with lightweight connectors that scale to hundreds of clusters without causing significant overhead. Key Topics Covered: • Why Crossplane's CRD-per-database model becomes resource-prohibitive at enterprise scale • How Klutch delivers Cloud Foundry-style developer self-service natively in Kubernetes • The architectural separation between the Klutch control plane and lightweight application cluster extensions • Why native Go operators are replacing Crossplane for certain backend integrations in Klutch • KROW (Kubernetes Resource Orchestrator) as an emerging integration technology alongside Klutch Read the full story & transcript at www.tfir.io #Kubernetes #PlatformEngineering #Crossplane #CloudNative #anynines #Klutch #DevOps #DataServices #OpenSource #DeveloperExperience
Kubernetes Self-Service Data Provisioning: What's Actually Missing | Julian Fischer, anynines
April 6, 2026 7:03 pm
Most platform teams reach for Crossplane when they need to manage data services across Kubernetes clusters — and most of them hit a wall. The moment you introduce dozens of databases, hundreds of tenants, and thousands of service bindings, the overhead becomes unsustainable and self-service for developers breaks down completely. Julian Fischer, CEO of anynines, explains exactly why Crossplane alone can't solve this problem — and how Klutch, the open source project at the core of a9s Hub, fills the gap with lightweight connectors that scale to hundreds of clusters without causing significant overhead. Key Topics Covered: • Why Crossplane's CRD-per-database model becomes resource-prohibitive at enterprise scale • How Klutch delivers Cloud Foundry-style developer self-service natively in Kubernetes • The architectural separation between the Klutch control plane and lightweight application cluster extensions • Why native Go operators are replacing Crossplane for certain backend integrations in Klutch • KROW (Kubernetes Resource Orchestrator) as an emerging integration technology alongside Klutch Read the full story & transcript at www.tfir.io #Kubernetes #PlatformEngineering #Crossplane #CloudNative #anynines #Klutch #DevOps #DataServices #OpenSource #DeveloperExperience
Kubernetes Self-Service Data Provisioning: What's Actually Missing | Julian Fischer, anynines
YouTube Video UEwxWXdxaGkyb3JOMGxxVERhNUxHUG1BdmdUMVRlWWhncy5FQjcyRDZBMEMyNTY3OTdE
Managing data services across hundreds of Kubernetes clusters, multiple AWS accounts, and on-premise environments has become one of the most painful unsolved problems in platform engineering. When you add compliance requirements like SOC 2, HIPAA, or FedRAMP, traditional tools simply don't hold up — and developer self-service breaks down completely. Julian Fischer, CEO & Founder of anynines, joins Swapnil Bhartiya on TFiR to discuss how the a9s Hub — built on the open source Klutch project — solves data service orchestration at enterprise scale across both Kubernetes and Cloud Foundry. From AWS multi-account architectures to on-premise sovereignty requirements, Julian explains why lightweight cluster extensions beat heavyweight control planes every time. Key Topics Covered: • Why Crossplane fails at scale for Kubernetes data service orchestration and how Klutch fills the gap • a9s Hub for AWS: orchestrating RDS, S3, and custom services across hundreds of AWS accounts using the Well-Architected Framework • a9s Hub for On-Prem: tackling data sovereignty and digital sovereignty requirements in OpenStack/VMware environments • The Klutch Network Connector Interface: exposing VMs and containers to Kubernetes application clusters seamlessly • How anynines enables uniform developer self-service (declare intent → get database) across multi-cloud and hybrid environments • AI workload runtimes: why Cloud Foundry and Kubernetes together form the foundation for stateful AI agents Read the full story & transcript at www.tfir.io #Kubernetes #PlatformEngineering #DataServices #CloudNative #anynines #Klutch #DevOps #MultiCloud #AWS #DigitalSovereignty
Kubernetes Data Services Are Broken at Scale — Here's the Fix | Julian Fischer, anynines
YouTube Video UEwxWXdxaGkyb3JOMGxxVERhNUxHUG1BdmdUMVRlWWhncy4wMEMyQTBBOTUwNTM0OUFG
GPU scarcity is the headline story, but what about AI workload orchestration? Julian Fischer, CEO of anynines, explains how his company approaches the AI infrastructure challenge. Instead of competing for scarce GPUs, anynines focuses on what they do best: wrapping complex, distributed open source projects in enterprise automation. Fischer reveals research into a9s Model—a product concept for on-demand LLM provisioning that applies their proven orchestration approach to AI toolchains. He also highlights vector databases as another natural fit for anynines' data service automation capabilities. For platform teams navigating AI adoption on Cloud Foundry and Kubernetes, this is a pragmatic view of how established orchestration patterns extend into the AI era without reinventing the wheel. Company URL: https://www.anynines.com Read the full story at www.tfir.io Become a Client/Sponsor: https://tfir.io/tfir-video-vip/ #AI #MachineLearning #LLM #VectorDatabases #anynines #CloudFoundry #Kubernetes #AIInfrastructure #Automation #KubeCon
AI Workloads on Cloud Foundry: a9s Model for On-Demand LLM Provisioning | Julian Fischer
YouTube Video UEwxWXdxaGkyb3JOMGxxVERhNUxHUG1BdmdUMVRlWWhncy41MzJFNEYxODEyNzA0QUUx