The factor your users actually carry
Biometric, hardware token, TOTP, SMS-based two-factor and email one-time codes are all supported, so the second factor can match the hardware your people already have.
OmniPriv is built on the assumption that something will eventually end up somewhere it should not. Every secret is encrypted, every component talks over mutual TLS, and every login can be made to prove more than a password.
Multi-factor authentication is enforced up front, a layered key hierarchy protects what is stored, and cryptographic hash-chaining protects what is recorded, so integrity holds even when the storage layer underneath is not trusted.
Identity is checked continuously rather than once. Nothing in the platform treats a successful login as proof of anything that happens afterwards.
Biometric, hardware token, TOTP, SMS-based two-factor and email one-time codes are all supported, so the second factor can match the hardware your people already have.
Keystroke rhythm, speed and pattern are compared against the identity's learned baseline. A drift triggers an MFA challenge automatically rather than being logged and ignored.
An administrator cannot reach credentials or approve requests outside their defined role boundaries, enforced at every access layer, not only in the interface.
Encryption only means something if it covers the whole path. Stored credentials, data in flight, and the internal traffic between OmniPriv's own components are each encrypted separately, so no single layer is the only thing standing between an attacker and a secret.
The harder question is not whether data is encrypted, but who holds the keys and whether anyone could change the record afterwards.
Audit records are held in tamper-proof storage with cryptographic audit-chain hashing, preserving integrity and non-repudiation, not merely retaining entries.
The platform itself contains no hard-coded credentials. Every secret it uses is vault-managed and auditable, including the ones that keep the platform running.
The installation key is generated at installation and must be stored externally and independently of the platform, then carried forward across upgrades and migrations.
Backups are fully encrypted with independent, secure key management, so a stolen backup is not a stolen credential store.
MFA at the front, layered keys behind it, mutual authentication between every component and hash-chained records underneath.
We will walk through the encryption model and the key custody arrangements against your environment.
Talk to the OmniPriv team about your identities, systems and AI agents.
Request a DemoBiometric factors, hardware tokens, TOTP authenticator apps, SMS-based two-factor and email one-time codes. Supporting several matters in practice, because a second factor only helps if the people who need it are willing and able to use it.
Keystroke dynamics are the rhythm, speed and pattern of how a person types. OmniPriv compares those characteristics against the identity's learned baseline at login, and a meaningful deviation triggers an MFA challenge automatically rather than being recorded and left alone.
Sensitive data is encrypted both at rest and in transit, with a layered key hierarchy (KEK, DEK, SEK) protecting the keys. Traffic between the platform's own components is mutually authenticated and encrypted, so there is no plaintext layer inside the stack.
It is the key generated when the platform is installed, and it must be stored externally and independently of OmniPriv itself. Keeping it separate is what stops a compromise of the platform's storage from also handing over the means to decrypt what is in it, and it has to be carried forward across upgrades and migrations.
No. OmniPriv contains zero hard-coded credentials, every secret it uses is vault-managed and auditable. That is a deliberate design constraint rather than a configuration option, because a hard-coded credential is one that no rotation policy can ever reach.