AI Systems
Part V covered the backend infrastructure every product needs. This Part covers the additional layer required specifically when a product incorporates a large language model — prompting, retrieval, agents, memory, and the evaluation and governance practices that keep an AI feature reliable rather than a novelty. These chapters assume the API and backend fundamentals from Part V.
Start Part VI — 6.1Prompt Engineering
The lowest-cost, highest-leverage lever available for improving what a language model produces.
6.2Context Engineering
A broader discipline than prompting alone — what information the model actually has access to, not just how it's asked.
6.3Retrieval-Augmented Generation (RAG)
How a language model answers questions about information it was never trained on.
6.4Embeddings & Vector Search
The mathematical trick that lets a computer judge whether two pieces of text mean similar things.
6.5AI Agents & Tool Calling
The shift from a model that only answers questions to one that can act on them.
6.6Agent Orchestration & MCP
What happens when a task needs more than one tool, one model, or one system to complete.
6.7Memory in AI Systems
Why an AI product that "remembers" the user is solving a harder problem than it looks like from the outside.
6.8AI Evaluation & Hallucination Reduction
"It seemed to work when I tried it" is not a testing strategy — this chapter is about what actually is.
6.9AI Safety & Governance
The set of guardrails that separate a genuinely useful AI feature from a liability waiting to surface.
6.10LLMOps & Model Selection
The operational discipline of running AI features in production, closing out this Part.