Founding thesis
Why the record has to be the product
The sentence everything hangs on.
Nobody sells governance of actions — the layer that scopes, gates, evidences, and attests an agent's individual act.
The existing "AI governance" category sells governance of models: inventories, policy documents, paper about the AI. The missing product is a control primitive for the work itself. wetink does not sell agents; agent vendors are a channel, not competition. It is the harness agents must wear, the mechanism that converts the risk officer's "no" into "yes, under these controls."
The upgrade hiding inside the constraint
Human review forced sampling: QC departments review a fraction of files1 because reperformance by hand costs too much. A governed machine reviewer changes the economics, and with them the control regime itself: structured rationale on the full population instead of a sample, each file with its own evidence pack. Better assurance and lower cost are the same product here, not a trade-off.
Why the record, specifically
A quality-control department's output is an evidence pack: home-lending review is the rare domain where the governance layer is not overhead, because the audit trail is the deliverable itself.
The product is not the automation.
It is the proof that the automation was accountable: who judged, against which version of which rule, citing which page of which document, at what cost, and whether anyone could have quietly changed the answer afterwards.
Prompts are guardrails for efficiency; the database is the guardrail for integrity.
The kernel treats the model as an untrusted client. Nothing it believes changes what the system of record accepts. Separation of duties, the control we expect examiners to test first, is enforced by the schema itself, not by a policy anyone has to remember to follow: the maker who does the work cannot seal it.
That distinctness is enforced between database actors today. Binding the checker to a verified human identity is trust-ladder rung three, and it is not yet built.
Inspect the enforcement model →
When proposed work becomes authority
The consequential moment is not the model's output. It is the instant a proposal becomes an independently authorized, attributable record — a distinct actor stands behind it, and the architecture makes that transition durable: proposed work → independent authority → accountable record. Every AI action carries that same structure, recorded in a form that outlives the meeting where it happened.
Current state. The sealing architecture runs today; the AI judgment layer that will operate inside it does not yet exist. See what runs today for the exact boundary. This site holds itself to that distinction as strictly as it holds any bank it hopes to work with.
Notes
If the thesis is right, inspect the architecture.