PUSH — capture
Every outreach message and meeting note lands in one place.
A capture habit your agent can read from — fed into the same Obsidian vault, not one more app to check separately.
By invitation · 4 sessions · 8 hours · $3,000/person
Your customer relationships and program status, remembered by your agent — not just you.
Four weeks, one session a week: customer outreach, meeting summaries, client knowledge, and program tracking — all captured once and recalled on request, built hands-on in Claude Chat, Cowork, and Code.
In practice: a CRM, without the complicated SaaS — your client history lives in your own Second Brain, not a $200/seat platform you have to configure and maintain.
The framework
WRITITATION: capture everything, retrieve anything.
Patent-pending, trademarked, and already proven: the same methodology behind a seven-figure FinOps engagement, now applied to customer outreach, meeting summaries, client knowledge, and program tracking through an Obsidian vault.
PULL — recall
Ask about a client or a program. Get an answer grounded in your own history.
Semantic recall over your Obsidian vault, built on Karpathy's LLM Knowledge Base pattern — not a generic search box, a connected memory.
How this is built
One workflow. Two layers.
You're not learning a tool. You're building a Second Brain that remembers your relationships, meetings, and programs, and hands all three back to you on request.
Tier 1 · Build layer
Claude Chat, Cowork & Code
Where you draft outreach, run Agent Skills, and wire up connectors — hands-on, across all three surfaces, so you know which one to reach for and when.
Tier 2 · Operating layer
Obsidian Second Brain
Persistent memory of every client and program, structured on Karpathy's LLM Knowledge Base pattern — where WRITITATION's capture and recall actually live, session after session.
What participants ship
A working Second Brain, built on your own relationships and programs.
A working agent stack
Claude Chat, Cowork, and Code set up and wired to your own email, Slack, and notes — not a sandbox demo.
A reusable Meeting Summary Skill
Turn any transcript or notes into a structured summary and action items, on demand, without doing it by hand again.
An Obsidian Second Brain
Client knowledge and program status, captured once and recalled by your agent — built on Karpathy's LLM Knowledge Base pattern.
Curriculum
Four sessions. One Second Brain that outlives the cohort.
8 hands-on hours over 4 weeks, one 2-hour session a week. Starts from zero: environment setup is folded into Session 1. Built directly on your own customer outreach, meetings, and programs — not a generic demo dataset.
Claude Chat, Cowork & Code — meet your agent stack
Week 1 · 2h
Claude ChatCowork & Code
You leave with: Your agent stack installed and running — Claude chat, Claude Cowork (Desktop), and Claude Code — plus a first personalized outreach message drafted for a real customer
Tools · What each surface is for (chat vs. Cowork vs. Code), environment setup, and your first standing instructions (CLAUDE.md)
Agent Skills — package your meeting summary habit
Week 2 · 2h
Agent SkillsMeeting summaries
You leave with: A custom Agent Skill that turns a raw meeting transcript or notes into a structured summary with action items, on demand
Tools · Skill authoring (the skill-creator pattern), invoking Skills from Claude Cowork and Code
Email & Slack integration — recall what you know about a client
Week 3 · 2h
Email & SlackClient knowledge
You leave with: A working connector that reads your inbox and Slack, and surfaces a client's history and open items before you respond — draft-then-approve, nothing sends unattended
Tools · Email and Slack connectors (MCP), draft-then-approve workflow design
Obsidian Second Brain — track every program, Karpathy-style
Week 4 · 2h
ObsidianProgram tracking
You leave with: An Obsidian vault, structured on Andrej Karpathy's LLM Knowledge Base pattern, wired to your agent — tracking project and program status across sessions, plus a rollout plan for your team
Tools · Obsidian vault structure, WRITITATION capture/recall applied through Obsidian, agent read/write access to the vault
Who this is for
Executives, builders, and operators responsible for relationships or program outcomes — and the team around them.
This cohort is built for executives, not engineers — no IT department, no procurement process, no developer required. If you can describe who you need to reach, what you discussed last time, and what you're tracking, you have everything this cohort assumes going in. Bring your Chief of Staff, Directors, and VPs: this runs deeper as a small team than as one person alone.
- CEOs & Founders — whose customer outreach and institutional memory are scattered across email, Slack, and their own head
- Chiefs of Staff, Directors & VPs — who carry the day-to-day follow-through on relationships and programs the CEO can't track alone, and who will actually run this system after the cohort ends
- Board members & advisors — who want a hands-on, non-technical way to understand what agentic AI actually does before recommending it to their own organizations
- Executives managing key accounts or programs — who lose time re-researching context and status that already exists somewhere in their own notes
- Business development & influencer/creator teams — managing a pipeline of prospects, partners, or followers across DMs, email, and spreadsheets that don't talk to each other
- Community builders — who need to remember every member's context and history without hiring a dedicated CRM admin to do it
Get started
Reserve a seat, or talk it through first.
Reserve a seat — $3,000/person
By invitation. Email directly to reserve one of the first cohort seats.
Open link →Corporate process review & consultation
Bringing this to your whole organization? Ask about a process review and consultation engagement beyond the cohort itself.
Open link →Talk to Ping about your org
Prefer a conversation first? Book time to discuss fit, timing, and your specific team's setup.
Open link →Prefer to go deeper first?
See the underlying case study and methodology this cohort is built on: 7 Figures of Impact.
Open link →Executive Agentic AI Cohort