Tag: Azure
-
Semantic Kernel, AutoGen, and Microsoft Agent Framework on Azure (Azure Gen AI Series, Part 21)
Semantic Kernel and AutoGen merged into Microsoft Agent Framework in 2026. Here is how the unified SDK works, what a migration costs, and where I would still wait.
-
Distributed Training on Azure ML with ND GPU Clusters (Azure Gen AI Series, Part 19)
How to spread one training run across Azure ND H100 clusters with PyTorch, NCCL, and InfiniBand, when adding nodes pays off, and how to keep scaling efficiency from collapsing.
-
Deploy Open Models on Azure Machine Learning with Managed Compute (Azure Gen AI Series, Part 18)
Open models on Azure Machine Learning run on managed compute, dedicated GPU VMs you rent by the hour instead of paying per token. Here is when that trade beats serverless, how to size the SKU, and where the community registry leaves you on your own.
-
Azure OpenAI Distillation and Stored Completions (Azure Gen AI Series, Part 17)
Capture production traffic with store=True, then distill a small Azure OpenAI model that answers like a flagship. The workflow, the real costs, and the traffic volume where it pays off.
-
Fine-Tuning Azure OpenAI, from SFT to DPO and RFT (Azure Gen AI Series, Part 16)
SFT, DPO, and RFT on Azure OpenAI: which models take which method, what the training and hosting actually cost, and how to read the loss curve before you deploy.
-
Azure Prompt Flow, Authoring to Evaluation to Deploy (Azure Gen AI Series, Part 15)
Prompt Flow now carries a retirement date. Here is how it works, when it still earns a place in your Azure pipeline, and where to build new instead.
-
Azure AI Content Safety, Prompt Shields to Groundedness (Azure Gen AI Series, Part 14)
Azure AI Content Safety scans prompts and responses for harm, jailbreaks, hallucinations, and copied material. Which API catches what, how the 0 to 7 severity scale drives a block decision, and where each check belongs in a request.
-
Azure AI Foundry Agent Service, from First Agent to Production (Azure Gen AI Series, Part 13)
Foundry Agent Service is the managed runtime in Azure AI Foundry that stores an agent, runs it against a conversation thread, and calls tools. Here is how the pieces fit, when to switch to standard setup, and where it breaks.
-
Azure AI Search and RAG, from Index to Answer (Azure Gen AI Series, Part 12)
How Azure AI Search grounds a model in your own documents: the ingestion pipeline, keyword and vector retrieval, RRF hybrid merging, and the semantic ranker that decides what the model actually reads.
-
Calling Azure OpenAI Models with REST, the SDKs, and the Responses API (Azure Gen AI Series, Part 11)
Azure gives you three ways to call a model, and they are not interchangeable. How Chat Completions, the Responses API, and raw REST differ, and when to pick each on the v1 path.
-
Azure OpenAI Entra ID Authentication, Managed Identities, and Encryption (Azure Gen AI Series, Part 10)
Move Azure OpenAI off static API keys to Microsoft Entra ID and managed identities, disable local auth without locking yourself out, and know when customer-managed keys are worth the cost.
Architect’s Toolkit
PJ’s Tools
VMware Cloud Foundation
- VCF Documentation
- VCF 9 Planning & Preparation Workbook
- VCF Bill of Materials (BoM)
- VMware Compatibility Guide
- VMware Interoperability Matrix
- VMware Configuration Maximums
- VMware Ports & Protocols
- VMware Hands-on Labs
- RVTools Download
Nutanix
AI & Cloud-Native Platform
- NVIDIA Build (Model Catalog)
- NVIDIA AI Enterprise Reference Architecture
- NVIDIA NIM Performance Benchmarking
- NVIDIA NGC Catalog
- NeMo Microservices Helm Chart
- Helm Charts Repository
- Hugging Face Models
Architecture & Design
About the Author
Dr Pranay Jha
Dr. Pranay Jha is a Cloud and AI Consultant with 18+ years of experience in hybrid cloud, virtualization, and enterprise infrastructure transformation. He specializes in VMware technologies, multi-cloud strategy, and Generative AI solutions. He holds a PhD in Computer Applications with research focused on Cloud and AI, has published multiple research papers, and has been a VMware vExpert since 2016 and a VMUG Community Leader.
You May Have Missed
DrJha