What we can prove today, stated plainly.
Trust is not something you are asked to extend on faith. Here is our current position, including the parts still in progress.
Compliance status
Our SOC 2 Type I assessment is in progress. It is not yet complete, and we will not describe it as though it were.
Human check, always
Every workflow keeps a human check at the exact point your own team already reviews that work today. That check does not move because a system is involved.
Audit access
Every evaluation, escalation and decision is logged in a form your own team can check independently, without needing Sagacian in the room to interpret it.
Documentation on request
A data processing agreement and penetration test report are available on request as part of any active conversation.
Where you should look for the detail.
We are also asked, reasonably, what actually decides an outcome and what happens when something goes wrong.
- What happens if it fails?
- Liability allocation on a failed guardrail is not yet finalised publicly. It is under active resolution and will be stated plainly once settled, rather than implied or left ambiguous.
- Can we take the workflow back if this does not work out?
- Contractual exit and handback terms are not yet finalised publicly. The same discipline applies: stated plainly once settled.
The architecture behind every decision.
AnchorPoint is what tests each action against your own rules and keeps the record your team can check, built into the engagement itself rather than a separate tool you take on. Read what it is and what it is not →