Category Knowledge
Technologies
The tools change. The categories persist. What follows is not a resume inventory but a map of capability areas I have worked across, with specific implementations as evidence of depth rather than boundaries.
Languages & Runtimes
C#, .NET Core, TypeScript, JavaScript, Python, SQL, KQL, and PowerShell
Each language chosen for specific runtime characteristics rather than brand loyalty.
Cloud Platforms
Azure, Azure Kubernetes Service, Functions, Logic Apps, Data Factory, Service Fabric, Cloud Services, and Terraform
The durable skill is cloud-native thinking: resource governance, cost modeling, network topology, and identity. These patterns transfer regardless of the console.
Container Orchestration
Kubernetes (AKS), Docker, and KEDA
The orchestration patterns of scheduling, service discovery, rolling updates, and autoscaling are consistent across environments. Autoscaling is proactive, not reactive: demand-driven capacity planning and queue-based scaling (KEDA) add capacity ahead of load, so throughput follows the work instead of catching up to it.
Distributed & Event-Driven Systems
Event-driven architecture, CQRS, Orleans actor framework, SignalR, and Service Bus (message brokers)
Architectural patterns that decouple services, separate read and write concerns, manage distributed state, and enable real-time communication.
Data Platforms
Azure SQL, MongoDB Atlas, Redis, QDrant, and ETL/ELT pipelines
The durable skill is modeling data for the access pattern. Connection strings are documentation, not architecture.
API & Integration
REST, gRPC, SignalR, and OAuth2
The pattern determines scalability, not the protocol. Secure API design protects every integration point.
Frontend
React, TypeScript, and JavaScript
The durable skill is component architecture: state management, rendering strategy, bundle budgets, and accessibility. The framework is an implementation detail.
AI & Machine Learning Systems
Claude Code, Ollama, RAG inference pipelines, Embedding pipelines, Vector databases, LLM optimization, and Azure AI Foundry
AI-assisted workflows act as a force multiplier, not just an accelerator. They compress low-value toil, widen the surface area a single engineer can validate and deliver, and redirect judgment toward the work that compounds — architecture, tradeoffs, and product insight. RAG pipelines keep that leverage honest, grounding outputs in verified domain data so speed never outruns correctness.
Quality & Testing
TDD, xUnit, TestContainers, and Automated CI validation
Systematic testing prevents regressions before they reach production and guides design decisions at every layer.
Technology is a means, not an identity. Every implementation I have used taught something the next one built on. The specific tools in each category will continue to evolve; the architectural thinking, the tradeoff evaluation, and the pattern recognition travel unchanged.