Platform Engineering & AI
Mark Silvester
I help platform engineering teams figure out how to operationalise AI. The infrastructure, developer experience, and delivery patterns that take AI from prototype to production.
Latest
-
QCon London 2026: Uncorking Queueing Bottlenecks with OpenTelemetry
Engineers from Gearset demonstrated how distributed tracing and Service Level Objectives resolve observability challenges in asynchronous systems by shifting focus from queue metrics to latency-based alerting.
-
Elastic Releases Version 9.3.0 with Enhanced AI Tools and OTel Support
Elastic 9.3.0 introduces GPU-accelerated vector indexing via NVIDIA cuVS integration, delivering performance improvements for retrieval-augmented generation systems, alongside deeper OpenTelemetry support and expanded security features.
-
Google Enhances Node Pool Auto-Creation Speed for GKE Clusters
Google Cloud has optimised GKE's node pool auto-creation to reduce provisioning time through more efficient request batching between the control plane and compute infrastructure, enabling faster node readiness across heterogeneous clusters.
-
Quesma Releases OTelBench to Evaluate OpenTelemetry Infrastructure and AI Performance
Quesma launched OTelBench, an open-source benchmarking suite that measures both the performance of OpenTelemetry pipelines under stress and how effectively AI agents can implement observability configurations.
-
Kubernetes Drives AI Expansion as Cultural Shift Becomes Critical
A CNCF report finds Kubernetes at 82% production adoption for AI workloads, but organisational and human factors — particularly siloed structures and lack of cross-functional collaboration — now represent the biggest obstacles to successful AI deployment.
-
Cedar Joins CNCF as a Sandbox Project
Cedar, an open-source authorisation policy language developed by AWS, has been accepted into the Cloud Native Computing Foundation as a Sandbox project, enabling developers to decouple access control from application code through policy-as-code principles.
-
Pulumi Adds Native Support for Terraform and HCL
Pulumi has announced native support for HashiCorp Terraform and HCL, enabling direct execution of Terraform code and state management within Pulumi Cloud.
-
Docker Kanvas Challenges Helm and Kustomize for Kubernetes Dominance
Docker has launched Kanvas, a new platform that automates conversion of Docker Compose files into Kubernetes deployment artifacts, positioning itself as a competitive alternative to Helm and Kustomize.