Security
Security is a property of the structure.
LensX holds your organization's most sensitive records: payroll, personnel, portfolios. It was built so that protecting them is not a policy layered on top, but a consequence of how the system stores facts.
Everything secure.
Data is encrypted in transit and at rest. Every workspace runs isolated, every session is authenticated, and nothing your organization loads into the graph is used to train models or shared beyond it.
Access, granted in levels.
Permissions are granular and explicit: per person, per team, per department, per record class. An analyst sees their scope; a director sees theirs. Levels are granted, adjusted, and revoked by you, and every grant is itself a recorded fact.
Every decision, traceable.
Any analysis, recommendation, or decision the system produces can be traced back to the exact input records it was computed from. When an auditor, a regulator, or your own board asks "why", the answer is rows, not weights.
Why this holds
Facts live in a deterministic graph, never in model weights.
AI reasons over the graph; it does not absorb it. That single design decision is what makes the rest true: outputs can cite their sources because the sources still exist as records, and an erasure request actually erases something, because there is a specific row to erase.
Our commitments
- Your data is never used to train models.
- Every access grant and change is logged.
- Erasure requests remove the records themselves.
- Security questions answered directly by the team.