ARCHITECTURE01
RAG vs. Fine-Tuning: Choosing the Right Architecture
A practical framework for deciding whether a requirement needs retrieval, prompt engineering, tool use, fine-tuning, or a combination.
FIELD NOTE / PREVIEWINSIGHTS / FIELD NOTES
Practical thinking for production AI and software.
Technical notes, architecture perspectives, implementation guidance, and evaluation practices for teams building intelligent systems.
INITIAL TOPICS / PREVIEW LIBRARY
QUALITY02
Testing AI Agents Beyond the Happy Path
How to evaluate tool use, permissions, retries, memory, escalation, ambiguity, and workflow completion.
FIELD NOTE / PREVIEWPROMPTS03
Prompt Engineering as Software Engineering
Why prompts need architecture, versioning, test cases, release processes, and monitoring.
FIELD NOTE / PREVIEWRAG04
What Makes an Enterprise Knowledge Assistant Trustworthy?
Retrieval quality, permissions, citations, data boundaries, evaluation, and operational ownership.
FIELD NOTE / PREVIEWDELIVERY05
From AI Prototype to Production System
The application, integration, testing, monitoring, and governance work required after the demonstration succeeds.
FIELD NOTE / PREVIEWEVALUATION06
Evaluating RAG Systems with More Than One Score
Evaluate retrieval relevance, context quality, groundedness, citations, completeness, and failure behavior.
FIELD NOTE / PREVIEW