Build Your AI Second Brain
End the cohort with a knowledge base that writes itself.
Your knowledge is scattered across 19 apps. None of it works for you.
Who this is for
Business owners, knowledge workers, and founders who want their notes, ideas, and decisions to compound.
- Business owners with 5+ years of notes scattered across apps
- Founders who think faster than they can capture
- Knowledge workers who keep re-Googling what they already learned
- Coaches and consultants whose IP lives in their head
What you ship
A living personal knowledge base in Obsidian, maintained by Claude Code, that compiles, cross-references, and answers from your own thinking.
- Obsidian vault structured as an LLM Wiki
- Claude Code agents writing pages from raw input
- Daily ingest ritual locked in
- 1,000+ pages compiled and cross-referenced
- Personal output workflows wired to the vault
You went from note-taker to knowledge architect. Your second brain runs while you live.
Week by week
1Vault foundations
Ships: Vault initialized + 5 first pages compiled
Vault foundations
Ships: Vault initialized + 5 first pages compiled
Workshops
- Install Obsidian + Claude Code
- Configure the vault structure (knowledge/, schema/, log)
- Karpathy's LLM Wiki pattern explained in plain language
- Your first /ingest-url command
Homework
- Pick 5 raw sources to ingest (notes, podcasts, articles)
- Run /ingest-url on each
- Read the auto-generated pages and flag what is wrong
Success criteria
- Obsidian + Claude Code installed and connected
- Vault folders created per the LLM Wiki schema
- 5 sources ingested and pages cross-linked
2Compile pipelines
Ships: 10 more pages, 4 templates, daily inbox ritual
Compile pipelines
Ships: 10 more pages, 4 templates, daily inbox ritual
Workshops
- Page templates: concept, project, person, source
- The /process-inbox command for daily fleeting thoughts
- How Claude decides which pages to touch
- Cross-reference rules
Homework
- Build 4 custom page templates for your domain
- Create your inbox.md and run /process-inbox daily
- Ingest 10 more sources, audit cross-references
Success criteria
- 4 page templates saved
- Inbox processed for 5 days running
- 10 new pages with intentional cross-links
3Daily rituals + scheduled agents
Ships: Daily ritual streak + 1 scheduled agent live
Daily rituals + scheduled agents
Ships: Daily ritual streak + 1 scheduled agent live
Workshops
- Morning ingest ritual (10 min)
- Evening reflection capture
- Scheduled Claude Code agents (cron-based)
- When the agent gets it wrong: lint and correct
- Office hours week: 15 min 1-on-1 with Rajiv for vault structure review
Homework
- Schedule 1 daily agent (e.g. summarize yesterday's notes)
- Run the daily ritual for 5 days straight
- Write a /lint-wiki report and fix 5 broken links
Success criteria
- Daily ritual hit 5 days in a row
- 1 scheduled agent running
- Lint report clean of broken links
4Domain knowledge graphs
Ships: Domain graph with 30+ sources + queryable
Domain knowledge graphs
Ships: Domain graph with 30+ sources + queryable
Workshops
- Pick a domain to deep-dive (your specialty)
- Build the domain's knowledge graph (concepts, people, sources)
- 30+ sources ingested in this domain
- Q&A: query your second brain like Wikipedia
Homework
- Pick your domain
- Ingest 30 sources in that domain
- Test queries: 'what does my second brain know about X'
Success criteria
- Domain picked + scoped
- 30 sources ingested
- 5 test queries with useful answers
5Cross-reference and lint
Ships: Lint clean + 20 relationships added + scheduled lint
Cross-reference and lint
Ships: Lint clean + 20 relationships added + scheduled lint
Workshops
- Cross-reference patterns that compound
- Detecting orphan pages and stale content
- Schema for relationships (person to project to company to outcome)
- Auto-lint on a schedule
Homework
- Run lint, fix every flagged issue
- Add relationships to 20 pages
- Schedule weekly lint agent
Success criteria
- Zero orphans flagged
- 20+ pages with explicit relationships
- Weekly lint scheduled
6Personal output workflows
Ships: 1 essay + 1 decision memo + maintenance plan
Personal output workflows
Ships: 1 essay + 1 decision memo + maintenance plan
Workshops
- The /write-from-brain workflow (essays, briefs, decisions)
- Decision memos sourced from your own thinking
- Capture-to-output loop
- Maintenance plan: how to keep this for 5 years
- Final cohort showcase + alumni community access
Homework
- Write 1 essay sourced from the vault
- Write 1 decision memo using vault context
- Build your 30-60-90 day maintenance plan
Success criteria
- 1 piece of original output published
- 1 decision memo logged
- Maintenance plan saved with calendar holds
Companion materials
Everything you need to ship, in-platform
22 homework tasks · weekly live sessions · capstone showcase
Preview the companionPick your cohort
Format
- Weekly live session: Tuesday 12pm ET, 90 min
- Homework: 3 to 4 hours per week
- Slack community open 24/7
- Office hours: 1:1 with Rajiv mid-cohort
- Lifetime access to recordings
- Final cohort showcase
Your instructor
Rajiv [last name]
Founder. Operator. Florida-based. Builds AI systems for himself and ships the patterns to you. Bio placeholder for Rajiv to personalize.
Real questions. Real answers.
Who is this cohort for?
How is this different from First AI OS?
What are the prerequisites?
What if I fall behind?
Payment or transfer policy?
Ready to ship Second Brain?
Rajiv reads every application. Reply within 48 hours.
Apply for Second Brain