Eyeballs on AISee the whole picture

Policy checks run on the device and inspect prompts locally. The prompt, file or pasted data never leaves the laptop or reaches us. We only record metadata such as which tool, which model, who and when. We never use customer data to train models.

When a prompt breaks a rule, Standpipe can block it, redact the sensitive part, or ask "Are you sure?" before it's sent. It understands meaning, so it catches private customer details or proprietary information that has no fixed format.

An enterprise AI plan only protects data inside that tool, and only while an employee is signed into the business account. Standpipe covers other AI tools and personal accounts on the device.

Yes. Most teams start with a free 7 to 14 day monitor-only trial on a small group. They review a report of the AI their people actually use, then turn on rules.

Initial deployment takes some IT work, including configuring MDM and updates. After setup, Standpipe runs quietly in the background, and employees do not need to open or configure it. Admins are notified if someone turns Standpipe off. OpsCompanion also offers managed services for Enterprise customers.

We maintain a SOC 2 program and pen test regularly. Every release is signed, and we support SSO, MFA, role-based access and audit logs. Our Trust Center covers sub-processors, incident response and our policies, and we'll give your security team access.