Experts engage
through the work.
This is the framework intended to govern practitioner work here: review on shipped artifacts, working sessions on real problems, approved field writing, and cohort teaching by people still doing the work. No expert has signed and none of those services is live.
No names appear here yet.
A practitioner becomes public on this page when an agreement is signed. Not when a conversation opens, not when someone is interested, not while terms are being negotiated. People we have approached and people we are talking to stay off the site entirely: no placeholder tiles, no silhouettes, no borrowed logos standing in for a commitment nobody has made yet.
The rule exists because an unverifiable roster is the specific thing wrong with this category. A name on a training site should be an obligation a reader can check — someone who agreed, in writing, to read shipped work and put their name on the verdict. Where that is missing, the roster is decoration, and decoration is easy to print.
So the count below reads zero. That is the true state of the founding roster today, and printing it is cheaper than explaining a name we had no right to use.
This panel is generated from the roster record, not written into the page. When a founding expert signs and agrees to be named, the name and the number appear here in the same deploy. Until then there is nothing to show, and nothing is shown.
Four ways signed practitioners can engage.
An activation framework, not a directory. Each mechanic is designed so expertise transfers through the work itself. None is represented as live before a practitioner signs and the operating lane is staffed.
Review is designed to start with shipped work.
When signed roster coverage activates, review begins with an artifact — a repo, an eval, a workflow, or a decision doc. The intended record is a written verdict on the work, inside the workspace.
The session format centers real problems.
Contracted working sessions would use member problems — a failing eval, a stalled agent, a rollout decision — with real screens and constraints. No session or cohort has run yet.
Field writing requires an approved source and agreement.
The editorial model can carry practitioner essays and field notes in The Signal after the practitioner signs and approves a content license. Today, The Signal is platform-authored.
Cohort teaching requires current practice.
A future practitioner-led cohort would use the same graph members train on. Current practice is the entry requirement to teach; activation also requires a signed engagement and a staffed cohort.
What a free account opens.
Membership is free and real — a working account, not a mailing list. It opens the current platform surfaces below. Practitioner review, paid cohorts, and enterprise delivery remain activation states; no paid cohort has run.
Platform-authored essays today; practitioner field notes require signed publishing agreements.
Member work, questions, and replies organized around the artifact.
The skill graph, operating tools, and artifact flow.
The working patterns behind the curriculum, in the open.
Artifact submission is built; human expert review activates only with signed roster coverage.
Share your context
The designed request begins with an artifact, error, or goal so a provisioned practitioner can review context before a call.
Get matched
Matching runs on discipline, not a generic queue. Activation depends on signed roster coverage.
Solve together
When booking activates, the format is a time-boxed video consult on the actual work, with the decision documented in the workspace.
One-on-one consults activate only when the founding roster can support them. No availability is listed before it exists.
The roster standard, plainly.
Current practice is the entry requirement. Names are announced as signatures land — never before. Until then, a roster seat is an ID and a discipline, not a persona.
Agentic systems
AI native engineering
FDE delivery operations
Healthcare transformation
Executive strategy
Prompt systems & evaluation
Cohort economics, only by contract.
The expert program supports cohort teaching and future consults. Revenue share and booking activate only under signed program terms.
Referral economics, only by contract.
Already writing or teaching about AI-native work? Bring your audience. Seat attribution and affiliate payouts activate only when a signed agreement defines the rate, scope, and reporting.
A preview of scheduling, video consults, and a context-rich workspace. Live booking is not active yet.
The curriculum graph is built. A contracted cohort can map practitioner teaching onto the same nodes learners train on.
A curriculum workbench and worked tool references are available today. External execution runtimes are not bundled into the platform.
Planned seat attribution, payouts, and learner routing. Commercial activation follows signed program terms.
You build, deploy, or advise on AI systems now — present tense. Having done it once is not the same as being able to teach it while it is still changing.
A repo, an eval, a rollout, a decision doc. Something with a result attached. A bio is not evidence.
Every activated review engagement must land as a written verdict on someone's shipped work. If that is not the job you want, this is not the roster.
Until a signature lands you are an ID and a discipline on this page. We do not list anyone as a draw before they have agreed to be one.
Apply to the founding roster.
Teaching a cohort, reviewing learner work, and delivering client engagements are three different commitments. Tell us which one you want and why that one — the answer decides who reads it next.
No queue position, no acceptance rate, no applicant count. No cohort has run and no expert has signed, so any number here would be decoration. Applications can be recorded now, but no response-time promise is active until practitioner operations are staffed.
Terms — revenue share, payouts, anything you would put your name on — are given to you in writing before anything is signed. We are not quoting a rate in public that we have not set.