Skip to main content
KeenSafe
Trust Center

Built to be operated inside the most regulated environments in the world.

KeenSafe runs inside Tier-1 banks, national grid operators, healthcare groups and federal-aligned defense suppliers. Every control on this page is enforced at the platform level — not described in a marketing PDF, not left to operator discretion.

Tenants in production
180+
Enterprise · MSSP · federal-aligned
Region patterns supported
14
Includes 4 sovereign-cloud overlays
Cryptography baseline
AES-256 · TLS 1.3
BYOK · HYOK · HSM-backed
Audit log retention
7y default
Configurable up to indefinite
Trust pillars

Twelve controls procurement teams ask about. Every one is platform-enforced.

We assembled this list from the actual questions our largest customers — and their auditors — ask during procurement. Each pillar links to the underlying enforcement; ask for the full implementation memo under NDA.

Tenant isolation

Every customer runs in a cryptographically isolated tenant — no shared application state, no shared keys, no shared evidence chain.

  • Per-tenant logical isolation enforced at API gateway, orchestrator, and storage tier
  • Per-tenant encryption keys derived from tenant-bound key material (KMS / HSM-backed)
  • Cross-tenant access requires a signed authorization manifest and is logged as a high-severity event
  • Tenant deletion is cryptographic — keys are destroyed first, blobs are zeroized within 24h
Encryption standards

TLS 1.3 everywhere in transit. AES-256-GCM at rest. Customer-managed keys supported across SaaS, single-tenant cloud, on-prem and air-gapped.

  • TLS 1.3 only — TLS 1.2 is deprecated for new tenants, with mTLS available for high-trust integrations
  • AES-256-GCM for at-rest data · KMS-fronted envelope encryption with per-object DEKs
  • BYOK and HYOK supported via AWS KMS, Azure Key Vault, GCP KMS and HSM (Thales / Entrust)
  • FIPS 140-3 validated cryptography modules available for federal-aligned deployments
Audit logging

Every operator action, every API call, every evidence-chain write — append-only, signed, and exportable to your SIEM.

  • Tamper-evident audit log written to append-only WORM storage, signed per-event
  • Real-time streaming to customer Splunk / Sentinel / Chronicle / S3 via native sink
  • Retention defaults to 7 years; configurable per-tenant up to indefinite
  • Independent operator-action visibility — KeenSafe staff actions are logged in the customer-visible audit feed
Region pinning & data residency

Tenant data, evidence, and processing stay in the region you select. Sovereign-cloud overlays for EU, UK, US, AU, and federal-aligned environments.

  • Region pinning enforced at orchestrator and storage layers — data never crosses region without explicit policy
  • Sovereign-cloud patterns supported: Bleu (FR), Delos (DE), GCC-High (US-Fed), IL5 / IL6 paths
  • Single-tenant deployment available in any AWS / Azure / GCP region
  • On-prem and air-gapped options for environments that cannot egress to a vendor cloud
Data retention & deletion

You decide what is retained, for how long, and what gets purged. Cryptographic deletion is the default.

  • Configurable retention per data class: scope inventory, attack-path evidence, audit logs, identity metadata
  • GDPR Art. 17 / CCPA right-to-delete supported within SLA — surfaced via tenant admin console
  • Cryptographic erasure available for high-sensitivity datasets — key destruction precedes data deletion
  • Backup and DR copies inherit the source retention class — no retention drift
RBAC & identity

Fine-grained roles, scoped tokens, short-lived credentials and SSO-everywhere — including for KeenSafe staff access.

  • Native SSO via SAML 2.0 and OIDC — Okta, Entra ID, Ping, Auth0, JumpCloud
  • Just-in-time provisioning + SCIM 2.0 for user lifecycle automation
  • Granular role catalog with 24 built-in roles; custom roles defined via policy DSL
  • Phishing-resistant MFA required for admin roles — FIDO2 / WebAuthn (passkeys), TOTP fallback
API security

Every API call is authenticated, signed, rate-limited and scope-bound. The platform exposes nothing more than it needs to.

  • OAuth 2.1 with mTLS-bound tokens · short-lived access tokens (15m default) with refresh rotation
  • Per-endpoint scope enforcement · token capabilities cannot exceed their grant
  • Rate limiting + DDoS protection at the edge · request signing on high-risk endpoints
  • OWASP API Security Top 10 mapped to platform controls and re-validated quarterly
Production-safe execution

No agents required for discovery. Validation is read-only by default. Exploit-grade payloads only execute against written-authorized scope.

  • Default-deny on any execution that could cause downtime — promoted to "verify only" mode
  • Production-safe validation primitives: read-only enumeration, topological proof-of-reach, sandboxed payload replay
  • OT and clinical exclusion lists honored at the orchestration layer — never sent to operators
  • Per-run sign-off workflow for any path that crosses a designated critical-asset boundary
Scope enforcement

Authorized scope is a first-class object. The platform refuses to run against any asset that is not explicitly in scope.

  • Written authorization manifest binds every engagement to allowed asset classes, regions, and operators
  • Per-asset scope checks at the orchestrator — out-of-scope assets are silently dropped from the run
  • Scope changes require multi-party authorization and are logged with reason codes
  • Customer-side scope-revocation API — pull a scope and every operator is denied within 60 seconds
Legal & privacy controls

DPA, SCCs, BAAs, and procurement collateral ready before the first call with your legal team.

  • Standard Contractual Clauses (EU SCCs 2021/914) executed at order time
  • HIPAA Business Associate Agreement available for US healthcare deployments
  • DPIA template + Transfer Impact Assessment ready to share under NDA
  • No customer data is used to train AI models · evidence chain is customer-scoped only
Compliance posture

Independently attested and audit-ready for the frameworks your auditors and regulators care about.

  • SOC 2 Type II · SOC 3 public report · refreshed annually
  • ISO/IEC 27001:2022 · ISO/IEC 27017 (cloud) · ISO/IEC 27018 (PII)
  • PCI DSS 4.0 SAQ-A (for tenants processing card data context only — KeenSafe never stores PAN)
  • HIPAA Security Rule, GDPR Art. 32, NERC CIP-014 evidence acceptance, IEC 62443-3-3 alignment
Secure architecture

Zero-trust internal architecture, hardware-key-bound operator access, signed evidence chain end-to-end.

  • Internal services use mTLS + SPIFFE workload identities — no shared secrets at the service boundary
  • Operator access brokered through a zero-trust gateway with FIDO2 hardware keys and session recording
  • Evidence is signed at point of generation; chain-of-custody is verifiable end-to-end
  • Quarterly independent penetration tests on the platform itself · summary letter available under NDA
Data lifecycle

What KeenSafe touches at every phase — and what it does with it.

Phase
Data class
Control
Scope authorization
Asset inventory · IP ranges · domains · OT asset classes · operator allow-list
Signed authorization manifest · stored in WORM evidence chain · revocable API
Discovery
Asset metadata · TLS certificates · DNS · open ports · identity inventory
Read-only primitives by default · all probes attributable to a single authorization manifest
Attack-path generation
Graph of relationships between assets, identities, and crown jewels
Generated inside tenant-bound compute · graph never leaves the tenant region
Safe proof-of-exploit
Reproducible per-step evidence · screenshots · network captures · response payloads
Crown-jewel boundary checks enforced · per-run sign-off for boundary-crossing actions
Reporting & retention
Executive reports · technical kill-chain · remediation packages · audit-grade evidence
Per-tenant key material · configurable retention · cryptographic deletion on request
Compliance posture

Independently attested. Audit-ready. Regulator-accepted.

We don't treat compliance as the goal — we treat it as a baseline. The frameworks below are the floor we operate on, not the ceiling we aim for.

SOC 2 Type II
Annual independent attestation
ISO/IEC 27001:2022
Information security management
ISO/IEC 27017
Cloud-security controls
ISO/IEC 27018
PII processing in the cloud
GDPR
EU SCCs + DPA + DPIA template
HIPAA
BAA available for US healthcare
NERC CIP-014
Evidence chain accepted by regulators
IEC 62443-3-3
OT / ICS security alignment
PCI DSS 4.0
SAQ-A scope (no PAN stored)
NIST CSF 2.0
Controls mapped end-to-end
CSA STAR Level 2
Cloud Security Alliance
TX-RAMP / StateRAMP
Roadmap · 2026
Reports and letters of attestation available under mutual NDA. Email [email protected] with your company domain to request the procurement pack.
Sub-processors

The vendors that handle your data — and the ones that don't.

The list below is exhaustive for SaaS deployments. Single-tenant and on-prem deployments may use a subset or none, depending on your region and operating model.

Vendor
Role
Region
Amazon Web Services
Primary SaaS hosting · KMS · S3 · EC2
us-east-1 · eu-west-1 · eu-central-1 · ap-southeast-2
Microsoft Azure
Single-tenant deployment · Key Vault · Sentinel sink
EU North · UK South · East US · Australia East
Google Cloud Platform
Single-tenant deployment · KMS · Chronicle sink
europe-west1 · us-central1 · australia-southeast1
Cloudflare
Edge WAF · DDoS · TLS termination
Global · region-pinned per tenant where required
Datadog
KeenSafe internal observability · no customer data
EU · US (region-pinned)
Okta
KeenSafe staff SSO + FIDO2 enforcement
US · EU
Incident response

What happens when something goes wrong.

Our IR commitments are operational, not aspirational. If a control on this page fails, you will hear about it from us, with enough detail and enough lead time to brief your own stakeholders.

P0 — confirmed compromise
Customer notification within 24 hours · written post-incident report within 72 hours · forensic artifacts available within 7 days.
P1 — material control failure
Customer notification within 72 hours · root-cause analysis within 14 days · re-validation of compensating controls before close.
P2 — partial control degradation
Customer notification within 5 business days · published change log entry · evidence of remediation in the next audit cycle.
P3 — informational
Surfaced via the customer trust portal · no individual customer notification required.
1.84M+
Validated attack paths
12.6M
Monitored assets
180+
Active integrations
38,400
Validation runs / week
14
Sovereign regions supported
74%
Faster MTTR · critical paths
Get Started

Request the full procurement pack

DPA · SCCs · SOC 2 Type II report · ISO 27001 certificate · DPIA template · pen-test summary letter · architecture diagram — sent within one business day to verified company domains.