Writing
Turning Obsidian Into a Queryable Second Brain with obsidian-llm-hub
A hands-on review of obsidian-llm-hub: RAG over your notes, LLM write access with guardrails, and why open source matters for a second brain you'll keep.
Your Meeting Transcripts Are a Training Corpus for Institutional Knowledge
LLMs make years of meeting transcripts and threads mineable. Why leaders should treat recorded, open communication as infrastructure — and the trade-offs.
Validation Is the New Bottleneck: Reviewing AI Output You Can't Read Line-by-Line
Parallel coding agents outpace line-by-line review. How engineering leaders should redesign validation around contracts, tests, invariants, and telemetry.
The Orchestrator Skill: One Conversation to Keep, Many Contexts to Burn
How a thin orchestration skill keeps one high-level AI conversation small while sub-agents absorb and discard the expensive context — plus the real trade-offs.
Build the Dashboard Before You Build the Feature
Why senior builders should treat internal developer control panels — snapshot, restore, seeding, flag toggles — as first-class infrastructure for a tight feedback loop.
The Hidden Tax of Being CTO, IC, and AI Orchestrator at Once
AI lets one leader operate across more of the stack at once. The real edge isn't capacity — it's deliberately managing attention. A frame for doing it well.
When We Stop Reading the Code
A forecast: when humans stop reading code and start reviewing contracts and telemetry, machines write optimizations no engineer would attempt by hand.
The chat box is not the answer
A lot of companies are adopting AI into their products right now. And a common theme I see is mimicking what the biggest success stories have done—OpenAI,…
Stop letting your AI coding assistant jump straight to code
Over the past few months, I've been experimenting with a simple shift in how I work with AI coding tools like Cursor and Claude Code: I ask them to plan b…
AI as a Force Multiplier, Not a Replacement
When integrating AI (LLM) systems into your business processes, there's one key principle to keep in mind: always keep a human in the loop.