ZEROSTAND
Self-hosted privileged access management

Access that ends on its own.

Short-lived, policy-checked access for people and workloads, with the credentials that must remain vaulted and rotated.

  1. RequestPick the safe or account and say how long.
  2. ApproveAn approver for the safe, or an Admin, decides.
  3. ConnectThrough Zerostand, which supplies the credential.
  4. ExpiresThe grant ends; open sessions are cut within 30 seconds.

How a grant works →

Challenges
Standing passwords. Shared secrets. Access nobody revokes.

Privileged access management was first built to store credentials for a few administrators. Today every service and pipeline needs privileged access, and a stored credential stays a liability until it is rotated.

30%of breaches analysed in the 2025 DBIR involved a third party, double the year before.1
22%of breaches began with credential abuse, the leading initial attack vector.2
64%of valid secrets found in public repositories in 2022 were still active four years later.3

Sources

  1. Verizon, 2025 Data Breach Investigations Report, 23 April 2025: "third-party involvement in breaches has doubled to 30%".
  2. Verizon, 2025 Data Breach Investigations Report, 23 April 2025: "credential abuse (22%) and exploitation of vulnerabilities (20%) continue to be the leading initial attack vectors".
  3. GitGuardian, The State of Secrets Sprawl 2026: tracking valid secrets detected in public repositories in 2022, "Four years later, 64% are still active and exploitable".
How it works

The zero-standing-privilege control plane

Every access, human or machine, follows the same three steps. Nothing reaches a target without proving who it is and passing policy; wherever the target allows, it receives a credential that will expire.

All identities
  • ◯People
  • ▣Workloads

Prove

  • IdP sign-in with MFA
  • Kubernetes and cloud attestation
  • OIDC from CI
  • Step-up MFA per target

Decide

  • Policy as code
  • Approval chains
  • Access windows
  • Ticket and reason where you require them

Lease

  • Short-lived credential
  • Brokered, recorded session
  • Expires on its own
  • Grants, approvals and sessions in the audit log
All targets
  • ▤Servers
  • ◫Databases
  • ⬡Kubernetes
  • ▭Web apps
See it in action

See the full access lifecycle

A 30-minute walkthrough with an engineer on our demo environment. Bring your use cases; we map them to your environment, and a proof of value then runs it on your own servers.

  • A person connecting to a server with an SSH certificate that expires
  • A one-time database login, approved in context, its statements logged, revoked at the hour
  • A workload attesting with its Kubernetes identity and receiving a credential
  • A vaulted account rotated, verified and reconciled while you watch
The request did not send. Please try again in a moment.

We use these details only to reply to you and arrange the demo. They reach us through our form provider and our email system. Please don't include passwords, hostnames or other sensitive details about your environment. By submitting, you agree to this use. Read our Privacy Notice for how we handle your data and your rights.

Request sent.

We will reply from a named engineer, usually within one working day, with a few times to choose from.

FAQ

Questions buyers ask first

Does Zerostand replace our password vault?

It removes standing credentials wherever a short-lived one can be issued instead, and vaults, rotates and verifies the accounts that must remain. You can start with one safe.

Do we need to install agents on every target?

No. Targets are reached over the protocols they already speak: SSH, WinRM, and the native database wire protocols. Linux servers need one sshd setting to trust the Zerostand certificate authority.

Where does it run?

As containers on infrastructure you control, on-premises or in your cloud. Zerostand is self-hosted: secrets are stored and used on infrastructure you control and are never sent to us. Data goes elsewhere only to destinations you configure, such as your SIEM, email server or ticketing system.

What happens if the control plane is down?

Run it as an active and a standby node on PostgreSQL. For the worst case, a sealed offline export of the break-glass accounts, encrypted to keys your security team holds.

How are machines and services handled?

As identities, under the same policy and audit as people. A workload attests with its Kubernetes or cloud identity and receives a credential that expires.