I’m Lukas, a software engineer and cloud/AI consultant focused on building robust, scalable systems with .NET and Azure.
I write practical guides on .NET, AI, agents, evaluation, observability, retrieval, testing, guardrails, and Azure.
Reference
Start here
Microsoft Agent Framework: Introduction
Where Agent Framework fits in the .NET AI stack, from agents and tools to MCP integration and workflows.
Read the introductionRAG with EF Core and pgvector
Build RAG retrieval in .NET with PostgreSQL, pgvector, and EF Core.
Read the articleObservability for Agents
Trace model calls, tool calls, token usage, and failures with OpenTelemetry, Aspire, and Application Insights.
Read the guideBuild and deploy an MCP server with .NET and Azure Container Apps
Build, containerize, deploy, and call a Streamable HTTP MCP server with the official .NET SDK.
Read the guideLatest posts
Tool Execution Needs Idempotency and Audit Trails
Implement atomic tool-execution claims in .NET with EF Core, separate dispatch intents from attempts, and recover lost responses through stored audit evidence.
Privacy Boundaries for AI Telemetry
Enforce AI telemetry privacy with versioned capture grants, atomic quotas across replicas, and evidence expiry that survives queues and evaluation copies.
Composing IChatClient Middleware: Why Order Matters
Compare cache, retry, and telemetry order around IChatClient with a fake provider, exact event traces, cancellation, and a streaming failure.
Evaluating Retrieval Quality in .NET
Evaluate real Qdrant results in .NET with a custom IEvaluator, labeled chunks, Precision@K, Recall@K, MRR, nDCG, and persisted HTML reports.
Assembling Context for the Model
Assemble model context from Qdrant results in .NET: deduplicate exact text, check a local token allowance, and trace citations to admitted evidence.