The CRM course was months ago. The briefing is tomorrow.

For airlines and training organizations. Between recurrent courses, each crew member gets short CRM and TEM action cards for the next simulator session, flight or debrief, written from their existing MBTI or Symbiotics ADAPT report.

In the product, admins see whether a report is loaded, not what it says.

Request a walkthrough

A 30-minute walkthrough of the crew and admin views, with sample data, so nothing needs uploading. We are inviting a small group of founding partners.

See what admins see

BriefingDebriefWhat admins see

Simulator session briefing

Blind synergy

Illustrative example. Hand-written sample, shortened, not output from a real report.

Challenge & response trigger

“Check that”

Expected response

Pause, restate the parameter, confirm or correct.

When to use

Any deviation from the briefed plan or an unverified parameter.

Synergy area

Structure meets early threat detection. One of you anchors the sequence, the other surfaces threats early.

Threat points & countermeasures

Terse communication under workload can hide intent. Countermeasure: Agree to state intent before action during high-workload phases.

Briefing do’s & don’ts

Do Name the top threat for this leg in one sentence.

Don’t Do not treat silence as agreement.

Organization admins see none of this.

Event debrief

Illustrative example. Hand-written sample, shortened, not output from a real report.

Actionable behavioural lesson

State intent before action when workload rises.

Cue

Any configuration change below 3,000 ft.

Next flight

Brief who calls the deviation and the exact phrase.

Mental model alignment

The crew held different assumptions about who owned the deviation call. Ownership of the go-around decision was implicit.

Stress impact on communication

Workload compressed communication into short, ambiguous phrases. Intent was not verbalised.

Organization admins see none of this.

Users & crews

Illustrative example. Not a real crew member.

Beside contact and team details, the user list shows one thing about a report: Loaded or Missing.

Loaded

No report, no profile, no routine outputs, no sharing pairs.

Read where the wall opens

How a report becomes a briefing

A report goes in.
A crew member uploads an MBTI Step I/II or Symbiotics ADAPT PDF, or an admin uploads a batch. We keep a behavioural profile drawn from it under a token, and the original PDF is not kept.
Crew describe what’s next.
Simulator session, flight, event debrief or training, with the aircraft, phase and scenario. Solo, or with a partner who approved first.
Action cards come out.
Anthropic’s Claude writes them from that profile and the session details. The briefing prompt requires operational language and bars diagnosing or ranking people, and the debrief prompt requires a blame-free tone.

Brief with a partner without swapping reports

Blind synergy starts with a request. The partner approves or declines for that routine type. Partners never see each other’s type, profile, report, sessions or reflections. The cards draw on both reports and go only to the person who asked, so they can still hint at how the partner tends to work. Requests expire after 14 days. An approval lasts 30 days and covers every card the requester generates for that routine type, with no notice to the partner.

Nothing runs in the cockpit

There is no live guide and nothing to open in flight. Crew prepare before the session and review after it. AI writes each card and no person reviews it, so treat it as a prompt for the briefing: your SOPs, your standard callouts and the commander’s authority come first. Bring a trainer to the walkthrough to test the sample phrases against your callouts.

What your admins see, and what they don’t

Profiles drawn from reports sit behind what the product calls the anonymity wall, under tokens such as PROF-0000A-A. The record behind each token also holds any name and email printed on the report, plus the file name. Treat the data as de-identified for your organization and identifiable to the super admin, not anonymous.

Illustrative example

	Crew member	Organization admin
Report	MBTI Step II	Report loaded: Yes
Trigger phrase	“Check that”	Not shown
Lesson	State intent before action when workload rises.	Not shown
Routine	Flight, saved to your history	Adds 1 to Weekly activity
Credits	Your CRM Credits and tier	Tier counts and average credits. If you run rewards, each redemption by name with its credit cost.
Sharing request	Approve or decline	Counts only. Who asked whom stays between crew members.

You can see by name who has a report loaded and who doesn’t, but the admin screens don’t open a report. Readiness tiers count CRM Credits, not proficiency. Do not use this for hiring, pairing or assessment.

Where the wall opens

Feedback coaching

An organization admin can get guidance for a named crew member, drawn from their report, including a short feedback reception profile such as “Receives feedback best with concrete examples and time to reflect.” It does not show the report, and the guidance is saved to that admin’s sessions under the crew member’s name. It is on by default and the crew app never mentions it, so the crew member is not asked or told. Any crew member can turn it off in their own settings with sharing level 0 (Private), but that setting describes only partners, so tell crew yourself.

Linking a token to a person

Batch uploads wait for the super admin, a Soft Skills Suit role that no one at your organization holds, to link each token to a person, by hand or by matching the email printed on the report. A crew member’s own upload links to them automatically. Every link is logged, in a log only the super admin can read.

Batch upload

An admin can upload report PDFs in bulk, up to 10 at a time. That admin holds the named files first, and the upload screen shows each file name beside its token. Each report stays unusable until the super admin links it to a person. Crew can upload their own report instead, which links it to them straight away.

Research export

The super admin can download one row per report, with type, four preference scores, gender and category, filterable by organization and without names. Every export is logged. Sharing level 0 (Private) does not remove anyone, and the download does not suppress small groups, so a rare combination in a small fleet could identify someone.

Request a walkthrough

Bring your crew or union representative. We’ll walk them through both sides.

Questions from training, safety and data protection

Can this reach management or line checks?

Reports and cards stay out of the admin views, which show report status by name, counts such as weekly sessions, tiers and average credits, and, if you run a rewards catalog, who redeemed what. Counts have no minimum group size, so in a small fleet they can point to individuals. The openings above are the exceptions, so read them first. We don’t sell this for checks, and we ask you not to use it for them.

Can crew say no?

We recommend opt-in with no consequence for declining, but the product doesn’t enforce that, so agree it with your crew representatives first. An admin batch upload can load a report without asking the crew member, and admins see by name who has none. A partner approves or declines each routine type, and an approval lasts 30 days. A single approval can’t be withdrawn early yet, but a crew member can set sharing level 0 (Private) in their own settings, which stops all partner use and feedback coaching for them.

Our crews don’t hold these reports. Can we still use it?

Not without a report. Cognitive Copilot reads MBTI Step I/II and Symbiotics ADAPT PDFs; it doesn’t run a test. If your organization holds the reports, for example from selection, agree with crew representatives how they are loaded first.

Does this meet our CRM requirement?

No. It has no regulatory approval. It is built around CRM and TEM concepts and is meant to reinforce your existing programme between sessions.

MBTI is contested. Why use it?

It is contested, so we use a report only as a starting point: how this person tends to communicate, decide and respond under stress. Crew who already hold one take no new test. Cards are behaviour prompts, not a verdict on anyone, and we don’t sell this for pairing or evaluation. We have no effectiveness data yet, so we agree measures with you before any trial.

Do tiers grade people?

No. Bronze, Silver, Gold and Platinum count CRM Credits earned, not proficiency.

Where does report data go?

The original PDF is not kept. Its text, including any name and email printed on it, goes to Anthropic’s Claude, server-side, to build a behavioural profile. We keep that profile, with the name, email and file name, under the token. Claude then writes each card from the profile or profiles and what crew type about the session, and that text is saved with the card. The product has no deletion function or set retention period yet, a crew member can’t replace or remove a report, and hosting terms aren’t published, so ask before anyone uploads.

What doesn’t it do yet?

It runs in a browser, on desktop or phone, in English, with a connection. Cards can take up to 90 seconds. It is aviation only. There is no app, offline use, SSO, integrations or email reminders yet. Reports must be text PDFs, not scans.

What does it cost, and what does a founding partner commit to?

A walkthrough commits you to nothing. There is no public price yet. We agree founding partner terms and measures with you before any trial.

See both sides for yourself

We are inviting a small group of training organizations and operators to become founding partners, with early access, a say in what we build next, and measures agreed with you before any trial. Cognitive Copilot is new, so we have no customers or results to quote, and the cards on this page are samples.

The walkthrough takes 30 minutes and covers the crew and admin views, with sample data.