Build your AI OS, lesson by lesson
Three tracks. A 13-lesson roadmap. Three complete lessons are live now, with a clear path for routing work between Gemini and Claude, building reusable prompt libraries, and automating the workflows you run every week.
Foundations
Build the mental model and tooling that everything else runs on. Most people skip this — and then wonder why their AI outputs feel inconsistent.
What an AI OS Is (and Why It's Different from Chatting)
The shift from tool-grabbing to system-building. What agents, tasks, and workflows actually mean — and why the distinction matters.
Gemini vs Claude: Route Work by Model Strength
A routing table for every task type. When to use Gemini's long-context intake engine vs Claude's precision output — with a copy-paste decision checklist.
The Anatomy of a Great Prompt
The five-block RCTFE framework that produces reliable outputs. Why 'be more specific' is bad advice, and what to do instead.
Build a Reusable Prompt Library
Your first compounding asset: a personal prompt library with 5 copy-paste templates covering the tasks you run every week.
Workflows
End-to-end systems for real knowledge work. Each workflow is manual-first, designed to be understood before it's automated.
Your First AI Workflow: Inbox Triage in 20 Minutes
A 3-prompt system that processes your inbox to three outputs: actions, deferrals, and delete candidates. Runs in 5 minutes daily.
Research & Summarization at Scale
Feed 100-page documents to Gemini, extract structured insights, and hand the synthesis to Claude for polished output.
The Writing & Editing Pipeline
A three-stage workflow: outline → draft → critique. Separate prompts for each stage produce better writing than one all-in-one prompt.
Meeting Notes → Action Items
Paste raw meeting transcripts and get structured decisions, owners, and deadlines — in a format your team will actually use.
The Weekly Review Workflow
A Sunday ritual: review last week's outputs, capture what worked, reprioritize for the week ahead. AI-augmented in 15 minutes.
Agents & Automation
Once you understand your workflows manually, you automate the right parts. This track covers agent design, chaining, and when NOT to automate.
Designing an Agent That Completes a Real Task
The anatomy of a working agent: model + instructions + context + tools. How to write a system prompt that produces consistent results.
Chaining Steps: Multi-Stage AI Pipelines
How to pass outputs from one model to the next. When to use structured intermediate formats vs natural language handoffs.
Guardrails & When NOT to Automate
The failure modes nobody talks about. Which decisions require human judgment, and how to build checkpoints into automated workflows.
Your Personal Daily Driver Setup
Pull everything together: a morning brief agent, a research pipeline, and an inbox triage that runs automatically. Your full AI OS.
Join the ongoing curriculum
All current lessons, every new monthly release, and a structured path from prompt basics to full agent automation. $19/mo. Cancel anytime.
Subscribe — $19/mo→Cancel anytime · No contracts · Monthly billing