Flagship Product · Privileged Access Management

Govern every privileged session - from request to audit

Enterprise privileged access management for cloud-native, hybrid and on-premises estates. Sessions held to the scope an approver authorised, AI that runs on local models inside your own deployment, and signed evidence your auditor can verify independently.

Five-stage access lifecycle

Request → Approve → Access → Monitor → Evidence

  1. 01Request & justifyUsers request access with a business justification.
  2. 02Approve & scopeMulti-step or parallel approval, with the approver reviewing and editing the command scope compiled from the justification.
  3. 03Just-in-time grantTime-boxed privilege with automatic revocation, and ephemeral accounts created on the target and removed when the grant expires.
  4. 04Brokered accessThe vault injects credentials at session start.
  5. 05Monitor & proveLive oversight, command capture and risk scoring during the session, then a signed evidence pack for the audit file.

17

Deployable microservices

78

Seeded platform definitions

8

Compliance frameworks

Local

AI runs in your deployment

Why legacy PAM falls short

Vault-centric tools were built for a different era. They record what an administrator did and let you find the violation afterwards - they never constrain the session itself.

  • ×Standing admin rights persist for days or indefinitely
  • ×Sessions are recorded and reviewed after the fact, never constrained
  • ×Thick clients and VPNs required for privileged connections
  • ×Cloud-hosted AI blocks regulated and sovereignty-constrained deployments
  • ×Evidence packs assembled by hand for every audit cycle

Built cloud-native for modern enterprises

ZenXPAM maps the same five-stage lifecycle to the frameworks you are audited against - with local-first AI, no standing privileges, and signed evidence packs your auditors can verify for themselves.

Explore capability docs

Differentiation

Four targets legacy PAM cannot hit

Engineered for organizations that cannot compromise on security, compliance, or operational efficiency.

Local-first AI

Risk explanation, intent compilation and natural-language policy run on a local model inside your own deployment, and a local provider is enforced in production - so privileged session data stays in your environment.

Intent-locked sessions

The justification on a request is compiled into an approved command scope and enforced at the target - preventively on Linux shell and database sessions, with recording and risk scoring elsewhere. Out-of-scope actions are blocked, not just logged.

Agentless browser access

RDP, SSH, VNC, web applications, Kubernetes consoles and published tools such as SSMS, DBeaver and PuTTY, delivered through the browser with credentials injected at connect time. Nothing to install, no VPN.

Evidence, not just logs

Hash-chained audit, signed evidence packs with a per-file manifest, and an offline verifier - so an auditor can check the evidence without having to trust the platform. Mapped to SOC 2, ISO 27001, PCI DSS, HIPAA, SOX, GDPR, NIST CSF and DORA.

Five-stage access lifecycle

Every privileged session follows the same governed path - request, approve, access, monitor and prove - with no gap for standing privilege to hide.

01

Request & justify

Users request access with a business justification. Attribute-based policy evaluates context on every request.

02

Approve & scope

Multi-step or parallel approval, with the approver reviewing and editing the command scope compiled from the justification.

03

Just-in-time grant

Time-boxed privilege with automatic revocation, and ephemeral accounts created on the target and removed when the grant expires.

04

Brokered access

The vault injects credentials at session start. Users never see, copy or cache a privileged password.

05

Monitor & prove

Live oversight, command capture and risk scoring during the session, then a signed evidence pack for the audit file.

Everything in the platform

Privileged access management, non-human identity governance, credential rotation, session recording and OCR search, compliance evidence, workflow automation and disaster recovery - every ZenXPAM capability, on one platform.

Real-time threat detection & risk scoring

  • Anomaly engine that evaluates behaviour during the live session, not after it
  • Per-user behavioural baselines and per-session profiles, with deviations flagged as they happen
  • Command classification: data exfiltration, lateral movement and privilege-escalation patterns
  • Live threat feed pushed to the console in real time, filtered by severity, user, session and time
  • Threat timeline with correlated sessions and linked audit records, plus an acknowledge and escalate workflow
  • Login, session and behavioural risk assessment, with an AI explanation of why something was scored risky and how to mitigate it
  • Adaptive MFA challenges driven by risk, with device fingerprinting and 30-day device trust
  • Automated response: in-app alert, email, step-up MFA, block the command, terminate the session or escalate
  • Anomaly list with filters, correlation and classification, and a security alert acknowledge-and-resolve workflow
  • OCR-based threat enrichment on graphical sessions, and scoring of the recorded session against what it was approved to do

Non-human identity governance

  • One place to own, rotate, deliver and retire service accounts, API keys, tokens, certificates and cloud credentials
  • Read-only inventory across AWS, Azure, GitHub and Google Cloud, with managed, system and unmanaged triage and owner assignment
  • Dormant-identity detection and expiry alerts, with a rotation-health overview for every identity
  • Scheduled safe rotation for AWS IAM keys and Azure client secrets, with preflight checks, grace window, verification before cutover and one-click rollback
  • Short-lived GitHub App installation tokens on demand, instead of long-lived personal access tokens
  • SSH key rotation on target authorized_keys files with pinned host keys
  • Workload secret delivery through TypeScript and .NET SDKs and the zenx-secret CLI, with previous values kept through the grace window
  • OAuth client-credentials and mTLS workload clients, with revocation through a durable outbox
  • ACME certificate renewal and Google Cloud service-account key rotation
  • Human and machine accounts kept apart, with an optional view of non-human rows on the user list

Credential rotation: passwords, keys, tokens & certificates

  • Passwords: Active Directory and LDAP over LDAPS, Windows through WinRM or local accounts, Linux and Unix over SSH, SQL Server, Oracle, PostgreSQL, MySQL and MariaDB, network devices and mainframe
  • Keys: AWS IAM access keys, Azure client secrets, SSH keys on target authorized_keys files with pinned host keys, and Google Cloud service-account keys
  • Tokens: short-lived GitHub App installation tokens on demand, Kubernetes service-account tokens, and OAuth client secrets for workloads
  • Certificates: mTLS client certificate registration, revocation and expiry alerts, ACME renewal with dns-01, and a documented SAML signing-certificate renewal procedure with metadata refresh
  • Policy-driven schedules from 1 to 3650 days, with on-demand rotation, a manual-approval mode and a minimum interval above the grace window
  • Generated passwords follow your password policy, and the new credential is verified by repeated sign-ins before the old one is retired
  • Grace window that keeps the previous credential usable while applications catch up, with one-click rollback and clean-up of what was left behind
  • Dependent systems updated automatically: Windows services and scheduled tasks that run as the account, and vault secrets that reference it
  • Preflight checks before a rotation starts, for example Azure secret-slot limits and AWS key counts
  • Every run has a correlation ID, a history ledger and a transaction log; passwords and keys never appear in logs or audit records
  • Failure alerts by email and in-app, an SLA monitor for overdue rotations, and policy-violation events
  • Secrets delivered to workloads through SDKs and the zenx-secret CLI, and encryption-key rotation with a runbook

Backup & disaster recovery

  • Sealed vault backups encrypted with a passphrase, opened and verified at creation so an unreadable backup never reaches your list
  • Created, verified, downloaded, uploaded and restored from Vault > Disaster Recovery in the console, or through the API
  • Restore checks the master password before touching anything, runs in a single transaction and rolls back on failure; it needs a typed confirmation and an administrator role
  • Per-tenant backups, so one tenant never sees another tenant's bundles
  • Offline recovery tool (zenxpam-recover) that can verify, list and extract credentials from a backup when the platform is gone
  • Two separate secrets protect a backup (bundle passphrase and vault master password), so losing one does not expose or destroy the other
  • Backups persisted on Kubernetes volumes, with every backup, verification and restore written to the audit trail
  • Database backup scripts with backup from the streaming replica, optional upload to cloud storage, a restore script and application rollback to a previous release tag
  • High availability on a five to eight VM topology with PostgreSQL replication, documented at near-zero data loss and a 5 to 15 minute recovery on async replication
  • Session recordings archived to local disk, NFS or SMB, or cloud object storage, with retention defaults and session storage settings
  • Site-level DR patterns for the same site, an alternate region and cloud, with quarterly drill guidance

Compliance & audit evidence

  • Framework management and scoring for SOC 2, ISO 27001, PCI DSS, HIPAA, SOX, GDPR, NIST CSF and DORA
  • Automated control checks with per-framework scores, drill-down into failing checks and deadline alerts
  • Regulatory requirement tracking with owners, status and evidence
  • Remediation plans with assigned tasks tracked to closure, and policy-to-control mapping
  • Certification campaigns (quarterly, semi-annual, annual) with approve, reject, bulk and delegate, and automatic escalation of overdue reviews
  • Violation detection: orphaned, dormant and over-privileged accounts, with severity and assignment
  • Signed evidence packs with a per-file manifest and an offline verifier
  • Hash-chained, tamper-evident audit trail with integrity and chain verification
  • Scheduled regulatory reports for SOX, HIPAA, SOC 2, GDPR and PCI DSS, exportable as CSV or JSON
  • Forensic timeline, event correlation, pattern detection and session reconstruction, with AI summaries and natural-language audit queries

Resource & account discovery

  • Network discovery by subnet or an explicit list of up to 256 addresses and hostnames, with hostname resolution and port probing
  • Remote inventory of each host through a central scan agent, SSH (Linux, with a Windows fallback) or WinRM (Windows): users, groups, services, installed applications and last user login
  • Partial results are tolerated, and collection warnings tell you exactly what could not be read
  • Active Directory and LDAP discovery that can run on a schedule as well as on demand
  • Review results, select what to bring in, and auto-provision resources and accounts with rules, platform selection and approval
  • Scheduled and automatic discovery, per-scan settings, one-click re-run with stored credentials that are never shown or logged, and cancel for a running scan
  • The scan agent is locked down: scoped authentication, a default-on network allow-list and bounded parallel probes
  • Every scan and re-run is written to the audit trail, with the credential itself never recorded

Non-human identity discovery (credential inventory)

  • Answers the question "which credentials exist in my cloud accounts that ZenxPAM does not manage?"
  • Reads AWS IAM access keys, Azure app-registration secrets and certificates, GitHub App installations and Google Cloud service-account keys
  • Strictly read-only: it never changes the cloud account and never reads, stores or shows a secret value
  • Opt-in per cloud account, with a scan interval from 60 minutes to daily, and Scan now with a cooldown
  • One finding per credential: provider, type, principal, age, last used, owner hint and expiry
  • Flags credentials that are stale (not used in 90 days), never used, without an owner or expiring
  • Triage into managed, unmanaged, system (ZenxPAM's own) and ignored, with a recorded reason for anything you hide
  • Credentials already under ZenxPAM are recognised automatically; any other one is brought under management in a click, with the provider details pre-filled
  • Plain error codes for scans, such as a missing IAM permission, each with a step-by-step "How to fix this" guide
  • Findings that vanish from the cloud account are kept for 90 days, then removed; large estates are capped at 20,000 findings per account
  • Permissions are separate: viewing, ignoring and linking, and turning sources on are three different rights

OCR indexing & session text search

  • On-premises Tesseract OCR reads the screen text of recorded RDP and graphical sessions, sampled every 2 seconds by default
  • Terminal sessions are indexed from the SSH and Telnet transcript, so shell and GUI sessions both become searchable
  • Every line of text is stored with its timestamp, in a full-text index, so a search result opens the player at the exact moment
  • Dedicated Recording Text Search page, with indexing status per recording and a heal path for recordings that were missed
  • Secrets visible on screen are redacted before they are indexed, including private-key text
  • Threat enrichment: deterministic keyword and pattern detectors raise risk events that seek straight to the moment in the recording
  • Intent review: screen text is scored after the session against the apps and commands the session was approved for
  • Optional live OCR on running sessions, off by default, with warn or terminate responses in the same Monitor, Warn and Strict modes
  • Runs entirely inside your deployment: no screenshots leave it, and no cloud vision service is used

Program & application restrictions

  • Allow and deny lists of applications compiled from the access request, and enforced for that one session
  • Baseline allow and deny lists per access profile, applied to every grant that uses it
  • RemoteApp-only sessions that publish just the approved programs instead of a full desktop
  • On full Windows desktops, AppLocker policy applied to the session host automatically on every grant, so a blocked program such as PuTTY is stopped by the operating system
  • A domain-level AppLocker policy created once per Active Directory forest and linked to the just-in-time user OU, with nothing to configure per host
  • Fail-closed option: if the restriction cannot be applied, the grant is blocked rather than issued unrestricted
  • Ephemeral identity and restrictions removed together when the grant expires, with a reconciliation worker for anything orphaned
  • Command allow and block patterns for shell and database sessions, with step-up MFA, block and terminate responses
  • Each step (identity created, policy applied, policy removed, failure) is a separate audit event
  • Uses WinRM to the session host and domain controller, with the Windows Application Identity service enforcing the policy on the host

Workflow & identity lifecycle

  • Visual workflow builder with approval, condition, delay, notification, API call, webhook, script and data steps
  • Multi-step and parallel approvals with auto-approval rules and workflow triggers
  • Template gallery with categories, tags, customisation and usage statistics
  • AI-assisted workflow generation and optimisation
  • Onboarding wizard and offboarding queue, single or bulk, with access revocation confirmation and welcome email
  • Workflow notification steps that deliver on their own, outside the original request
  • Ticket-gated approvals: ServiceNow and Jira state checked live, failing closed if the ITSM cannot confirm
  • Instance tracking with progress, history and statistics

Access & sessions

  • Agentless browser access to RDP, SSH, VNC, web applications and Kubernetes consoles - nothing to install, no VPN
  • Published tools such as SSMS, DBeaver, PuTTY, pgAdmin, Chrome and Edge, with credentials injected at connect time
  • Database sessions and mainframe or telnet terminals
  • Standard, just-in-time and emergency (break-glass) access in one workflow
  • Ephemeral accounts created on the target on grant and removed when it expires
  • Jump servers, on-premises site tunnels and secure tunnelling with password injection
  • Pre-connect actions on the target, with cleanup on disconnect
  • Live spectate, take-control, pause, resume and terminate, with a multi-session monitoring grid
  • Session recording with playback, keystroke and command capture, and OCR text search over graphical sessions

Intent-locked enforcement

  • The written justification becomes an approver-editable command scope
  • Monitor, Warn and Strict modes, with step-up MFA, block and terminate thresholds
  • Deterministic, sub-second pattern matching on the command path, with the AI used only for grey-zone scoring
  • Preventive enforcement on Linux shell and database sessions; recording, reconstruction and risk scoring on other transports
  • A trace from justification to approved scope to commands to deviations, exported for audit

Private AI

  • Runs on a local model inside your deployment, with no outbound connectivity needed
  • Natural-language policy authoring, with simulation against historical access and coverage-gap analysis
  • Risk explanation, audit summarisation and compliance question answering
  • Notification composition, workflow suggestions and infrastructure health narration
  • AI chat with semantic search over the platform

Policy & access control

  • Attribute-based policy evaluated on every request, with a visual builder
  • Policy versioning, templates, simulation and testing before enforcement
  • Policy analytics: effectiveness, violation trends, coverage and recommendations
  • Tenant-scoped roles and permissions, entitlement groups and access reports
  • Access analytics and a custom report builder with saved templates, scheduling and export

Vault, rotation & discovery

  • Encrypted vault (AES-256) with checkout leases, versioning, history and policy-compliant password generation
  • Rotation for Windows, Linux, Active Directory, databases and network devices, with schedules, history ledger and failure alerts
  • Account discovery across Active Directory and the network, with auto-provisioning rules and scheduled scans
  • Encryption-key rotation with a runbook, and credential disaster-recovery guidance

Identity & integrations

  • SSO through SAML and OIDC, LDAP and Active Directory sync, and SCIM 2.0 provisioning
  • MFA with TOTP, SMS, email, biometric and WebAuthn / FIDO2 keys, with admin enforcement and compliance reporting
  • ServiceNow and Jira integration with ticket write-back on approve, deny and session events
  • SIEM streaming as JSON, CEF or LEEF, syslog over TLS or S3, per-sink event filters, dead-letter queue and replay
  • Notifications by email, SMS, push, Slack and Microsoft Teams, with templates
  • Published OpenAPI, SDKs for TypeScript, .NET and Python, a Terraform provider and a Kubernetes operator

Platform & deployment

  • Multi-tenant, with signed per-tenant entitlements and a cross-tenant console
  • Cloud-native microservices on Docker or Kubernetes with Helm, and no appliance
  • On-premises and sovereign hosting, with offline installation scripts and a guided installer
  • High availability on a five to eight VM topology with PostgreSQL replication and a credential disaster-recovery runbook
  • OpenTelemetry tracing and metrics, health endpoints, tenant-aware rate limiting and API quotas
  • CodeQL, Trivy and Gitleaks scanning in the delivery pipeline

Deployyour way

Cloud-native & hybrid

Containerised microservices on Docker or Kubernetes - no appliance, and no Windows Server or SQL Server estate to license.

On-premises & sovereign

Runs entirely in your own datacentre or private cloud, with AI features that keep working without outbound connectivity.

Multi-tenant & managed

One deployment serves many tenants, each with a signed entitlement, under a single cross-tenant console - built for managed PAM delivery.

Compliance evidence

Signed, framework-mapped evidence for audits and continuous assurance - mapped to SOC 2, ISO 27001, PCI DSS, HIPAA, SOX, GDPR, NIST CSF and DORA, the frameworks your auditors already use.

ISO 27001
SOC 2
HIPAA
PCI DSS
GDPR
  • Evidence packs signed and verifiable offline
  • Hash-chained, tamper-evident audit trail
  • Clause-level control mapping across eight frameworks
  • FIDO2 / WebAuthn MFA with adaptive step-up

ZenXPAM - Frequently Asked Questions

Ready to modernize privileged access?

Schedule a demo with our team or download the product brochure and capability documents for your security and procurement review.