Blog
Build notes, engineering decisions, and ideas in progress.
I write about designing practical AI systems, working with LLMs, RAG and agents, and building knowledge workflows that support clearer thinking. The focus is on what worked, what failed, and what I would change.
Writing in progress
The first build note is taking shape.
New articles are being developed around practical AI engineering, project decisions, and Personal Knowledge Management. Nothing here is presented as published work before it is ready.
pipeline
What I Learned Designing a Five-Agent Curriculum Pipeline
A transparent look at the architecture decisions, orchestration challenges, quality gates, and trade-offs behind the paused Quest & Crossfire Curriculum Studio project.
The project is paused, but its engineering lessons remain useful. This article will distinguish completed work, design decisions, unresolved problems, and what I would change next.
Editorial roadmap
Planned articles
These are working topics, not publication claims. Titles and scope may evolve as the underlying notes and examples are developed.
When Should You Use an AI Agent—and When Should You Not?
A practical framework for choosing between agents, deterministic workflows, and conventional automation.
AI EngineeringDesigning Notes for Retrieval: Where PKM Meets RAG
How note structure, metadata, context, and source quality influence retrieval and grounded AI responses.
AI + PKMA Practical Evaluation Checklist for LLM Features
A repeatable way to examine quality, grounding, failure cases, latency, cost, and real user value.
AI EngineeringEditorial approach
What these notes will prioritize
The goal is not to publish generic AI news or repeat documentation. Each article should reveal useful reasoning that a practitioner can examine, adapt, or challenge.
Decisions over demos
Why a design was chosen, which alternatives were considered, and what constraints shaped the result.
Failures and trade-offs
What did not work, what remains uncertain, and where a simpler approach may be more responsible.
Reusable lessons
Patterns, checklists, and questions that remain useful beyond one project or tool.
Continue exploring
Choose structured learning or inspect the work.
The Learning Hub organizes evergreen paths for AI Engineering and PKM. The Portfolio shows how these ideas are applied in projects and engineering decisions.