Author Archives: Allen Rodriguez

Microsoft AI-103: Agent Retry Policies Retries are one of the easiest ways to make an agent system look more reliable while quietly making it less safe. A transient 429 from a model deployment may deserve another attempt. A malformed tool request does not. A timeout after a payment API call is especially dangerous because the […]

Microsoft AI-103: Agent Conversation State Conversation state is what lets an agent behave as though the second turn belongs to the first. It sounds simple until the application has to decide which messages belong together, how long they should persist, what happens when two requests arrive at once, and whether the conversation is allowed to […]

Microsoft AI-103: A2A Protocols for Enterprise Agents Agent-to-agent communication becomes difficult when every team invents its own invocation contract. One agent exposes a REST endpoint, another expects a function call, a third returns a proprietary task object, and none of them describe capabilities in the same way. The Agent2Agent protocol is meant to reduce that […]

Microsoft’s agent platform is no longer one product with one runtime. The same platform boundaries are central to Microsoft AI-103 preparation, where agent design is best understood as architecture rather than as isolated product features. Teams can choose how much orchestration Microsoft manages and how much they own. Microsoft Foundry Agent Service can run prompt […]

“custom commands” is still a useful way to think about repeatable Claude Code workflows, but the current product model is broader: Anthropic has merged custom commands into skills. The useful takeaway is how this behavior changes the workload boundary. Existing Markdown files under .claude/commands/ continue to work, while new reusable workflows are generally better created […]

Microsoft AI-103: API Management for AI Gateways When a company has one AI application, it is easy to connect that application directly to a model endpoint. The design changes when dozens of applications, agents, teams, and tenants share model deployments. They need a common place to authenticate callers, allocate capacity, observe consumption, enforce policy, route […]

AZ-500 Retirement and the SC-500 Cloud Security Transition AZ-500 retired on August 31, 2026. Microsoft’s direct replacement is SC-500, the Cloud and AI Security Engineer Associate exam that became available on July 21, 2026. For anyone preparing now, the practical conclusion is simple: AZ-500 is a legacy exam, while SC-500 is the current Microsoft cloud-security […]

Security posture management creates a long list of issues: exposed services, missing protections, vulnerable software, risky permissions, stale identities, weak network boundaries, and configuration drift. Operators often see this as a prioritization problem. An attacker sees it as a path-building problem. The attacker does not need every weakness to matter; only a sequence that reaches […]

Identity protection and privileged access controls are often strong on paper. Multifactor authentication is enabled, privileged roles are listed, risky users are monitored, and administrators know that least privilege matters. Designs still fail because standing access remains too broad, emergency accounts become ordinary accounts, risk signals are ignored, approvals become automatic, or operational pressure encourages […]

Azure network security is often drawn as a collection of products and policy objects: network security groups, Azure Firewall, web application firewalls, DDoS protection, private endpoints, routing, and segmentation. The diagram can look correct while the deployed environment is still weak. The gap usually comes from how people request rules, interpret exceptions, troubleshoot outages, inherit […]

Azure Key Vault is often introduced as the place where secrets, keys, and certificates should live. That is correct, but incomplete. The real design problem is how applications and people reach those objects, how privileged changes are approved, how network access is constrained, how rotation occurs, and how teams keep the vault from becoming either […]

Microsoft Defender for Cloud architecture is easy to reduce to a dashboard, a score, and a list of recommendations. That view misses the harder design problem. The service sits across resource inventory, policy, posture management, workload protection, permissions, multicloud connections, security operations, and remediation ownership. Choices in one area change what the platform can see […]

Microsoft AI-900: Retirement and the Current AI-901 Fundamentals Exam AI-900 was retired on June 30, 2026. That makes one point unambiguous for anyone preparing now: AI-900 is a legacy exam, not the current Azure AI Fundamentals destination. The current exam is AI-901, which Microsoft updated for a fundamentals audience that now needs both conceptual AI […]

Microsoft AI-901: Responsible AI Principles and Accountability Responsible AI becomes difficult when a team has to make a real decision before it has perfect evidence. A model is about to be piloted, data coverage is uneven, some user groups are underrepresented, and the business wants a launch date. The responsible choice is not produced by […]

Microsoft AI-901: NLP and Text Analysis — Entities, Sentiment and Summaries Natural language processing becomes useful only when text is converted into a decision that the surrounding system can understand and govern. The difficult part is rarely the definition of sentiment analysis, entity recognition, summarization, or classification. It is knowing how raw language becomes structured […]