Real access to your systems, under your control
org.run does real work in your real systems. So every action is scoped, logged and reversible, and anything high-stakes waits for a person.
rule: Support replies run on their own
Principles
What we hold to
An operator needs more trust than an advisor. These four ideas shape everything we build.
Control
You decide what org.run can touch and where it must stop. Permissions mirror your org, down to the system, action and dollar amount.
Visibility
Every action has a receipt. You can see what it did, when, in which tool and why.
Data protection
Your data is encrypted, kept in your own systems where possible and never used to train models.
Reversibility
Writes are recorded with what changed, so a wrong action can be rolled back instead of cleaned up by hand.
Inside every run
The safeguards in every run
Encryption
TLS in transit and AES-256 at rest. Credentials are sealed in a vault and never shown to the model.
Isolation
Every run happens in its own isolated sandbox. One task cannot see or affect another.
Least privilege
Credentials are scoped to the task at hand. Permissions mirror your org, so org.run never has more reach than the person it acts for.
Where it stops
Auto, ask or never
You sort every kind of action into one of three tiers. org.run follows the line you draw.
Done on its own
Routine, low-risk work it handles end to end, then logs.
Example
Reply to a routine support ticket
Waits for a person
It prepares the work, then stops for your sign-off before anything goes out.
Example
Issue a refund over $500
Off limits
Actions it will not take, whoever asks and however it is phrased.
Example
Delete customer data
Data governance
Your data stays yours
Never used for training
Your data is never used to train models, ours or anyone else's.
Stays in your systems
org.run works in the tools you already run. Your records stay where they live today.
Retention you control
You set how long anything is kept. Disconnect a tool and its tokens are revoked and its data deleted on your schedule.
Running a security review?
Talk to us
Bring your security team and your questions. We will walk through how org.run handles access, data and approvals in your setup.
Request our security questionnaire
We will send our answers on access, data handling and approvals. If you have your own vendor questionnaire, send that and we will fill it in.
Early access
Put org.run to work, safely
Start read-only. Draw the lines. Let it earn the rest.
Scoped, auditable and reversible from day one. Your data stays yours.
