byrxj · People Operations

The AI era is the People Operations era.

Products commoditise inside two release cycles. Categories rename themselves between board meetings. The only thing that compounds is the institution underneath, and the only function whose mandate is to design the institution is People Operations. Most companies have asked it to operate one instead.

If the institution is the next great moat, the people who design the institution are the moat itself.

How to use the triptych

Where the three diagnostics disagree is where they are most useful.

Step 1 · Run RADAR for the AI-shape read.

Does the company actually have an AI operating model, or is it running pilot theatre? RADAR's pathology output names exactly where the shape is missing.

Step 2 · Run FGK for the People Operations architecture read.

Tier-calibrated against Native / Compounder / Fortress / Spectator. The bottleneck axis names the binding constraint on the institution's ability to compound talent.

Step 3 · Run CoG for the strategy-structure coherence read.

Does the stated strategy match the observed gravity? Drift is the honest axis; counter-strategy promotions and story-shape mismatch are the load-bearing tells.

The diagnostic signal is the disagreement.

RADAR high, FGK low = the AI shape exists but the people machine cannot staff it. FGK high, CoG low = the People Ops architecture is sound but the institution is orbiting the wrong strategy. RADAR low, CoG high = the strategy and gravity match, but neither has actually moved toward AI yet. Read the gaps. They are where the work is.

Player-tier calibration

“Great” is not uniform.

FGK and CoG are tier-calibrated. Same raw score reads differently depending on which game the company is playing. The first move is not architectural. It is admitting which tier you are in.

Native
AI is the product.

Anthropic, OpenAI, Cursor. No version of the company survives without frontier model access.

Compounder
Existing moat, AI as multiplier.

Google, Adobe, Salesforce. AI is leverage on a real pre-existing business.

Fortress
Regulated, deliberate.

Banks, defense, healthcare. Slow by design because the cost of error is asymmetric.

Spectator
Talking, not yet operating.

The most common honest answer for established companies in 2026. Naming it is the prerequisite for moving out of it.

See where your institution stands.

Start with FGK. Four minutes. Tier-calibrated. A verdict you can quote in a board meeting.

Take the FGK Diagnostic