Sample chapters
Six excerpts from the work
Six selected excerpts give a view of retrieval, hallucination, agents, security, testing and human-AI interaction. Each stands on its own and needs no registration.
Available
The excerpts
Chapter 14 · Volume 1
Retrieval, Chunking & Hybrid Search
Why vector search is only the beginning of good retrieval.
- Retrieval
- Chunking
- Dense/Sparse
- Hybrid Search
- RRF
Chapter 16 · Volume 1
Understanding hallucination
Why plausibility is not truth.
- Factuality
- Grounding
- Consistency
- Hallucination
Chapter 18 · Volume 1
From prompt to agent
How reasoning, acting and observing become a controlled agent loop.
- ReAct
- Tool Use
- Agent Runtime
- Guards
- Human-in-the-Loop
Chapter 27 · Volume 2
Prompt injection
Why content carries no authority, and how attacks propagate through agentic systems.
- Prompt Injection
- Indirect Injection
- Agent Security
- Trust Boundaries
Chapter 33 · Volume 2
AI testing
How non-deterministic systems become testable with the right test oracles.
- Testing
- Evaluation
- Test Oracles
- Determinism
- Production
Chapter 35 · Volume 2
Human-AI UX and trust
Why good AI products calibrate trust rather than maximise it.
- Human-AI UX
- Calibrated Trust
- Approval
- Product Safety
All six files are excerpts, not complete chapters, and they open as PDFs in a new tab. The work is written in German, so the sample PDFs are German; there are no English excerpts.
The work
Where the excerpts come from
The samples are taken from AI Engineering, a three-volume work on the architecture, governance and operation of production AI systems.
View the work