Blog
Engineering deep dives
without the marketing
We write about what actually works: RAG architecture on .NET, the pitfalls of bringing AI into corporate processes, legacy modernisation. Only what we have verified in projects.
12 deep dives
Filter by topic ↓Contextum: the document pipeline to Markdown for AI agents
AnetLab Contextum: a five-stage pipeline between an agent and an LLM — Markdown conversion, context compression, idempotent delivery and an honest loss registry.
RAG evaluation on .NET: Recall@5, MRR and the release gate
How to measure RAG quality on .NET: a Golden Set of 50–100 questions, computing Recall@5 and MRR, measuring faithfulness and eval-as-a-release-gate in CI.
Agentic workflows in enterprise: patterns and anti-patterns
When agents pay off in corporate processes and when a deterministic workflow is enough: architecture, typical failures and a rollout checklist.
Engineering leadership in AI products: evaluation, roles, delivery
What changes in a tech lead’s job with AI products: eval as an engineering practice, ownership of the answer, team topology and the CI/CD/CE cycle.
Enterprise AI architecture: four decisions that decide everything
A practical guide to the architecture of corporate AI systems: RAG vs fine-tuning, agents vs workflows, quality evaluation and governance.
Production RAG: a 50-point checklist for production
A checklist for taking RAG to production: security, quality evaluation, observability, cost control and content lifecycle governance.
Taking an AI PoC to production: 90 days by phases
Why AI pilots do not reach production and how to fix it: Golden Set, eval pipeline, integration with the corporate context, LLMOps and a readiness checklist.
Modernising legacy .NET: WCF, AppDomains and the AI bridge
What really breaks when moving from .NET Framework to modern .NET: BinaryFormatter, AppDomains, WCF, YARP, Strangler Fig and the AI Bridge pattern.
MCP for enterprise: protocol, security and boundaries
How to embed MCP in the corporate environment: the protocol primitives, the security model, schema bloat, call auditing and the choice between MCP, REST and Kafka.
.NET and Azure for AI systems: a reference architecture
How to assemble a production-ready AI system on .NET and Azure: async pipeline, RAG, Polly, observability and cost control — with working code.
Building a RAG pipeline on .NET: a practical walkthrough
A step-by-step build of a production-ready RAG on .NET 10 — from text extraction to the evaluation pipeline. With the real decisions, not abstractions.
RAG on .NET: an architecture that survives production
Why naive RAG breaks on real data, and how we build a .NET pipeline: chunking, hybrid search, evaluation and cost control.
Topics
Five directions we write about. Pick a topic and the list filters itself.
- 01AI/LLM in productionRAG, agents, quality evaluation, cost control and running AI systems on .NET.9
- 02.NET platform and legacyFrom .NET Framework to modern .NET: WCF, AppDomains, Strangler Fig, YARP and engineering practice.6
- 03Architecture and patternsReference architectures, trade-offs, the limits of applicability and typical mistakes.3
- 04MCP and integrationsStandardising tool access, contracts with external systems, security and audit.2
- 05Engineering leadershipRoles in AI teams, quality criteria, the delivery cycle and process maturity.2
Have a task that needs a second head?
Describe it and we will answer on the merits — no pre-sales, no slide decks.