Guru Bala

Just read an excellent deep-dive article written by our own AI Cloud Solution Architect Jian Sheng Ng on implementing Agentic AI frameworks using Azure Kubernetes Service (AKS) with a robust multi-tenancy approach. This is exactly the kind of architectural thinking we need as AI agents become more prevalent in enterprise environments. Key highlights that caught my attention: 🔒 Security-First Design: Complete tenant isolation at multiple layers - compute, data, and identity - eliminating cross-tenant data leakage risks 💰 Cost Optimization: Shared infrastructure with dedicated compute per tenant, allowing for optimal resource allocation and VM type selection based on tenant profiles 🏗️ Enterprise-Grade Architecture: • Control plane isolation through Kubernetes namespaces • Compute isolation via dedicated node pools with taints/tolerations • Identity isolation using Azure managed identities • Data isolation through RBAC and ABAC on storage accounts 🤖 Semantic Kernel Integration: Practical implementation showing how to deploy AI agents that can plan, reason, and execute function calls while maintaining strict security boundaries The solution elegantly addresses the “noisy neighbor” problem while providing the flexibility enterprises need for AI workloads at scale. The use of WorkloadIdentityCredential for secure, temporary credential delegation is particularly well-executed. For anyone working on enterprise AI deployments or multi-tenant architectures, this is a must-read case study with actual implementation details. Check out the full technical walkthrough: https://lnkd.in/gCmr6vYM #AzureKubernetes #AgenticAI #MultiTenancy #SemanticKernel #CloudArchitecture #AzureAI #Kubernetes #EnterpriseAI Tony Spelde Johnny Tian Jian Sheng Ng Sebastian Henze Hazirah Hasnan Pradeep Menon

33

Agentic AI is powerful—and unpredictable. That’s why observability (and traceability) is non-negotiable. This week our SigNoz team was invited to Nutanix HQ in Bay Area to present “Taming Agentic AI with Observability.” Loved the conversations with the cloud-native community and seeing how fast teams are moving Agentic AI from demos to production. Why this matters: - LLMs are non-deterministic → you need end-to-end traces to see what actually happened. - Trace agent loops, tool calls, vector store lookups, and external API hops—along with latency, errors, retries, and token costs. - Use OpenTelemetry for a standard, vendor neutral instrumentation model. If you’re building Agentic AI, check out what’s possible with SigNoz + OpenTelemetry. Links for docs below or feel free to DM me 🎥 Clip from the talk below featuring Goutham Karthi Big thanks to the Nutanix team for hosting! Also it was great meeting Deepak Goel Shreyans Mulkutkar Devyansh Mittal and many other folks in the community

55

1 Comment