applicable.ai
Coming Soon Get Notified Sign In
Product

What it does, step by step.

You describe your institution once. It does the reading from there.

Eight steps

Each step shows its work.

1

Describe your institution.

Charter, regulator, states, products, entities, size. We can draft the profile from public sources. Drafted facts stay marked unverified until you confirm them, and some will be wrong.

Screenshot of the profile facts for the fictional Demo Community Bank: charter type, regulator, FDIC status, total assets, branch footprint and product lines.
Demo Community Bank's profile, fact by fact.
2

It reads each rule.

Each duty is read once for who it binds and what triggers it, carve-outs included. That reading is kept between runs.

3

It decides, and shows why.

It checks those conditions against your profile. Numbers like asset size are compared in code. The confidence says how much rested on interpretation.

Screenshot of one applicability decision for the fictional Demo Community Bank: the citation, the section text, the reasoning, and a “What would settle this?” question with Yes, No and Ask someone buttons.
One decision: the rule text, the reasoning and an open question.
4

When it can't tell, it asks.

A missing fact becomes a question. Answer it yourself, or send it to whoever knows. They answer from a link and don't need an account.

Screenshot of the email a colleague receives: who is asking, what the question is about, and Yes and No buttons.
The email a colleague gets. They answer from a link.
5

It watches the text.

When a section's text changes, the change is recorded and sorted by whether it matters. Decisions that read the old text wait for you. Proposed rules get a stage bar and a recommended response.

Continuously monitor regulatory developments. Coming soon

Screenshot of one horizon event's overview: a stage stepper, the AI rationale, an impact summary and key facts.
A horizon item's stages: evidenced, implied, or not yet reached.
6

It links a change to your controls.

Identifies impacted compliance artifacts and recommends changes or follow-up actions. Coming soon

Today, for a change or a horizon item, it suggests the duties and controls it may touch, with the evidence, and you accept or reject each.

7

Requirements become controls.

The Coverage Agent works on controls and requirement mappings. Duties that ask for the same thing are grouped. For each group it reuses a control you have, proposes a change to one that nearly fits, or drafts a new one. The proposal waits for you.

Screenshot of the controls library with one control open, showing its owner, cadence, testing procedure and regulatory basis.
A control with its owner, cadence, testing and regulatory basis.
8

You keep the record.

Each accept and override is saved with who did it and when. Your applicability decisions export to CSV.

Export the controls your team has accepted, and their requirement mappings, as CSV for import into your GRC. Coming soon

What it doesn't do.

It connects to a GRC by CSV export only; there's no API. One profile per account, and that profile can hold more than one legal entity. It doesn't attest or grade anything, and nothing counts as reviewed until a person acts on it.

Coming Soon

Be the first to know.

We're building a new AI-powered approach to regulatory compliance. Sign up and we'll tell you when access becomes available.

We'll email you when access opens, and nothing else.