Knowledge
Articles & insights
Short, practical explanations of the building blocks I use in client projects: RAG, MCP, workflow automation, local models, Office 365 integration and more. No buzzword bingo – just enough depth so you understand what you’re buying.
What is LLM and what it isn't
LLMs in Plain English: The Basics for Business Workflows
5–7 min read · For IT & business stakeholders
From statical to agentic workflows
Why Large Language Models change business process automation and how LLM based agents modernize traditional workflows.
5–7 min read · For IT & business stakeholders
Enterprise IT assets become available to LLM based intelligent agents
Large Language Models are capable to query databases, read files from shared folders, or call APIs without hard coded programmatic logic.
5–7 min read · For IT & business stakeholders
How RAG actually works – and when it makes sense
Retrieval-Augmented Generation explained in normal business language: what it is, when it’s overkill, and where it can really help.
5–7 min read · For IT & business stakeholders
Evaluation and testing of Retrieval-Augmented Generation systems
An overview of commonly used approaches for evaluating RAG systems, including retrieval metrics, answer groundedness, LLM-based evaluators, and the role of curated and synthetic evaluation datasets.
6–8 min read · Technical overview
What does a simple AI automation really cost?
A realistic breakdown of model usage costs for email and ticket automation — with concrete numbers, assumptions, and provider comparisons.
3–5 min read · For business & technical decision makers
MCP: connecting AI agents safely to your existing systems
A pragmatic look at the Model Context Protocol and how it can structure integrations between AI assistants and your existing APIs, databases and tools.
4–6 min read · For technical decision makers