Know what leaves.
And where it goes.
Follow the document journey from employee to AI tool, with file details, timestamps, and evidence in one view.
Your team is moving with AI. Know which tools they use, what they share, and where sensitive data goes.
AI moves fast.
Bring its footprint into focus.
Bring people, documents, and AI destinations together. Get the context to support good decisions, without slowing good work.
Follow the document journey from employee to AI tool, with file details, timestamps, and evidence in one view.
Surface personal information and confidential content. Understand the finding, its source, and its confidence.
Review findings, tune policies, and document the response. Start with visibility; add controls as coverage matures.
Connect activity to an employee, department, and managed device.
Identify the shared file and its personal or confidential information.
See the AI tool and the evidence supporting the observed action.
Review the risk, apply your policy, and keep a clear decision trail.
Build on the identity and information labels your organization already uses. Our Microsoft 365 integration is planned to connect the dots between people, documents, and AI activity.
Integration is in development. Coverage depends on tenant permissions, licensing, and supported file types.
Metadata and redacted findings first. Full document retention is not the default.
Keep confirmed actions, inferred activity, and uninspected content distinct.
Role-based access, tenant separation, and audit trails are core launch requirements.
Start with one team, a clear set of AI tools, and the data you care about most. Measure coverage and finding quality before expanding.
Choose your team and endpoints
Define sensitive data and AI tools
Validate evidence and detection
Agree rollout and commercial terms
Early product stage. Pilot availability and pricing will be agreed before deployment.
You can explore an interactive product demo today. The production service, Microsoft 365 connector, and managed endpoint agent are being developed in stages. The demo uses fictional data and does not monitor your device.
ChatGPT, Claude, and Kimi are our initial targets. Coverage will be validated separately for browser uploads, pasted content, desktop applications, and API traffic. A supported AI service does not imply coverage of every interaction.
No. Microsoft 365 provides identity, file, and label context. Reliable evidence of a third-party AI upload also requires a supported browser or endpoint observation. The platform is designed to show the source and confidence of each finding.
The planned default is to classify locally and send metadata plus redacted finding summaries. Any exception for retaining evidence content will require an explicit administrator policy and retention limits.
Not yet. The Windows agent is a later development phase. It requires an enrolled device, administrator deployment, signed components, compatibility testing, and transparent organizational policies before use.