michaelnguyen.net

Skills & Proficiencies

A quick snapshot of what I’m strong in, plus what I’m actively upskilling.

Currently upskilling

Skills I’m actively investing in right now (and using on this site):

AI Agents
Orchestration patterns, tool routing, guardrails, and agentic workflows.
Example: paddle shop + automation workflows
Terraform
IaC patterns, remote state, environments, and CI-driven deploys.
Example: SWA infra + deployment pipelines
Azure Static Web Apps
Multi-app deployments, subdomains, Functions API, and CDN edge behavior.
Example: main site + playground deployed independently

Core strengths

Microsoft infrastructure

  • Windows Server, IIS, .NET (Framework + Core)
  • Active Directory (operations, hygiene, troubleshooting)
  • PowerShell automation (5.1 + 7)

Cloud & platform

  • Azure (core services, identity/RBAC, networking fundamentals)
  • AWS (working knowledge; learned on cloud-native teams)
  • Hybrid/on-prem integration patterns

Observability

  • Monitoring design for workflows & business-impact views
  • Pragmatic alerting: reduce noise, improve actionability
  • OpenTelemetry concepts (signals, context propagation, OTLP)

Operations mindset

  • Incident response & troubleshooting
  • Cost awareness (right-sizing & decommissioning)
  • Build visibility that helps during outages

Observability viewpoints

Tools come and go. I care about signal quality, correlation, and the ability for humans to reason about production under stress.

My take

  • AppDynamics is strong for business impact in legacy/monolithic systems, but it rewards intentional instrumentation — not “monitor everything.”
  • Datadog is the closest you’ll get to a true single pane of glass with excellent correlation, but you need cultural readiness to manage cost.
  • LogicMonitor works well in MSP-style environments where you need to onboard diverse infrastructure quickly (agentless).
  • New Relic deserves mention for its generous free tier (you’re being monitored on it now).
OpenTelemetry

I see OpenTelemetry as a force for good: standardization around OTLP, schemas, and semantics is a net positive. But portability is overrated — observability is foremost a data problem. A top-tier presentation layer that helps humans reason about systems matters more than avoiding vendor lock-in.