Early access
SOPs and policies that stay tied to the rules underneath them
Policy Automation
- 9 weeks to 12 days
- median SOP update cycle in early access
- 100%
- of clauses traceable to an obligation
- 0
- SOPs of unknown currency status
An SOP library that is not linked to the obligations underneath it is a set of documents nobody can prove is current.
The problem
What this module exists to fix
SOPs drift from the rules
Annual review cycles mean an SOP can be wrong for eleven months before anyone looks at it. Inspectors increasingly ask when you knew and how fast you responded.
Nobody knows what a change breaks
When a rule amends, finding the affected clauses across 400 SOPs, policies and work instructions across a dozen sites is a manual search exercise, done badly, under time pressure.
Drafting is the bottleneck
Even once the gap is identified, getting text drafted, reviewed by legal, approved by a committee and attested by staff takes months.
Capabilities
What it does
Clause-to-obligation mapping
Ingest your existing SOP and policy estate and map each clause to the obligations it satisfies — automatically proposed, human confirmed.
Break detection
When an obligation changes, every dependent clause is flagged with the nature of the break: no longer sufficient, now over-restrictive, newly required, or now redundant.
AI-assisted drafting and redlining
Generate proposed replacement text in your own drafting conventions and tone, delivered as a tracked-changes redline with the source provision cited alongside every edit.
Full policy lifecycle
Draft, review, legal sign-off, committee approval, publication, staff attestation, exception management and scheduled review — with an immutable audit trail.
Hierarchy management
Model the real structure: policy, standard, SOP, work instruction, control description — and cascade a change down the hierarchy with a single review.
Coverage assurance
The inverse view: which obligations have no policy or SOP coverage at all, ranked by materiality and personal liability exposure.
How it works
Step by step
- 01
Ingest
Import the SOP and policy estate from SharePoint, Confluence, or a file share, preserving structure and version history.
- 02
Map
Clauses are matched to obligations with confidence scores and confirmed by policy owners.
- 03
Monitor
Obligation changes propagate to dependent clauses and raise a review task with the break explained.
- 04
Draft
Proposed redlines are generated, cited and routed through your existing approval path.
- 05
Attest
Published changes trigger targeted attestation to exactly the populations affected, per site.
Questions
The things people actually ask
Including the ones where the answer is a limitation.
- Does the AI publish policy text on its own?
- No, and it is not designed to. It produces a drafted redline with citations. Approval remains with your named policy owner and whatever committee your governance requires. The audit trail records who approved what and when.
- What formats can you ingest?
- Word, PDF, HTML, Confluence and SharePoint pages, and Markdown. Structure and numbering are preserved so clause references remain meaningful.
- When can we get access?
- Policy Automation is in early access with a limited cohort now, with general availability targeted for the second half of 2026. Existing customers get priority access.
See Policy Automation against your own footprint
A two-week scoped pilot on your real entities and jurisdictions. You compare the output against what your team found in the same period.