BÁEZ LABS
Sovereign AI for Executives
A hands-on 4-hour session where you install, configure, and run your own AI agents — on your own machine, under your own control. Most AI advice assumes your data can leave the building. Yours can’t.
C-Level executives (CEO, COO, CTO, CFO...), SVPs, VPs, and Directors who want to understand what an “AI agent” actually is before their team builds one — or before they sign a contract for one.
The CEO seat gets the most value: no role has more repetitive, judgment-light work orbiting it — executive reporting, meeting follow-through, market research, agenda management — that a local agent can absorb.
3–5 executives per session, minimum 3 to run. Small by design — with hardware-dependent local setup, more than 5 people means someone is stuck troubleshooting instead of learning.
A mandatory 15-minute async pre-flight, two days before, keeps a broken driver — or a slow model download — from eating into the room’s four hours.
The 4 Hours
Framework + Install
“Agent = Model + Harness” — why local matters: privacy, cost, control. Install Ollama, pull the LLM, configure Open WebUI to chat locally.
Deploy Two Agents, Live
Inbox Concierge + Draft Desk — already coded, pulled from repo. Each executive ends the block with two agents running for real, against their own data.
Hand Off Three More
Decision Memory, Boardroom Prep, Document/Contract Analyzer — pre-built templates to take home and deploy on your own. Stuck? That’s the door to a paid 1:1.
Close
Leave with five agents in a repo — two running for real, three ready to try — and an invitation to a paid 1:1 follow-up.
What You Need Before You Arrive
This founding cohort is exclusively Apple Silicon. Windows, Linux, and AMD aren’t eligible for this seat — reach out and we’ll set you up with a private 1:1 session or the in-house corporate format instead, which run on any hardware.
Apple Silicon M2 Pro or better
32GB unified minimum — 48GB+ recommended
60GB free
Updated to the latest stable release
Not eligible: 8–16GB Macs, Intel Macs, or any non-Apple-Silicon machine. Spec verification (a screenshot of “About This Mac”) is required before your seat is confirmed — if your machine doesn’t qualify, it’s a full refund or a free hold on the next cohort.
Two more things before the session: install Ollama and pull a small test model (ollama run gemma3:4b) 2–3 days out, and complete the mandatory 15-minute pre-flight call, where we also pre-download the session’s real models so the room’s wifi isn’t the bottleneck. And bring a real example of what you want to automate — a week of email exports, a meeting doc, a report. That’s what separates this from a YouTube tutorial: you leave with an agent running on your own data, not a demo.
Individual Seat — Founding Cohort
First 10 seats at founding pricing, per market. Regular pricing applies after the founding cohort fills.
The Same Content, for Your Whole Team
One private session, one invoice. Priced by session — not by seat — and tiered by headcount, since larger groups need more hands-on support during setup.
Available add-ons: a paid 1:1 follow-up session per executive, 30 days of async support, and a second deep-dive session at 60% of the base price.
Reserve a Seat in the Founding Cohort.
Your seat is held from the moment you reserve — we’ll reach out to coordinate the date and pre-flight.