Kliper
Sign in Book a demo
Kliper
TRUST CENTER

Every claim, with its evidence.

Kliper is built by compliance assessors, so we hold the platform to the evidence standard we apply to our customers. This page is the one place to evaluate our security and privacy posture, and every claim links to the artifact behind it.

Independent verification.

Third-party scans of our edge security configuration. Every one is public and can be re-run against us by anyone reading this page.

What we commit to.

Status

Real-time service status

Live uptime monitoring for the app, the client portal, the API and authentication, with public incident history.

status.kliper.dev

Architecture

Security architecture

How we protect evidence: a SHA-256 hash of every upload, malware scanning with ClamAV and a VirusTotal hash lookup, strict tenant isolation, and guardrails on Cortex.

Security overview · docs.kliper.dev/security-and-ai-trust

Cortex

AI in PCI assessments

Cortex, our AI assistant, is built to the PCI SSC “Integrating AI in PCI Assessments” guidance. It is a tool for the assessor, never a replacement for their judgment.

  • The assessor decides. Cortex proposes findings and drafts; every change it makes asks for confirmation first, and the QSA makes the compliance determination.
  • No fabrication. It answers from the assessment’s own data and cited PCI guidance. When there is no data, it says so.
  • Traceable. Answers cite real PCI DSS requirement numbers and quote saved values verbatim.
  • Sensitive data protected. Card numbers, IP addresses and other personal data are redacted before any AI call.
  • No training. Cortex runs on OpenAI and Anthropic models through their APIs, whose terms exclude API data from model training. A self-hosted model profile is available where prompts must stay on your own infrastructure.
Subprocessors

Subprocessors

The authoritative list of every third party that processes customer data: what each one receives, the region it processes in, and a link to its DPA. Customers under a DPA get 30 days’ notice before a new subprocessor is added.

docs.kliper.dev/subprocessors

Disclosure

Vulnerability disclosure

We follow RFC 9116. Found something? Email security@kliper.dev. We acknowledge within 2 business days and do not pursue legal action against good-faith researchers.

/.well-known/security.txt

Recovery

Data protection & recovery

All customer data is encrypted in transit (TLS 1.2+) and at rest. The production database ships its write-ahead log continuously to off-site object storage for point-in-time recovery, alongside nightly full dumps kept for 30 days.

  • RPO target: 5 minutes or less. In practice the write-ahead log ships at least every 60 seconds.
  • RTO target: 4 hours or less.
  • Restore verification: weekly, by an automated drill that restores the backup and checks it. A drill that fails to run raises an alert.

Customers on a paid plan can request our disaster-recovery one-pager and the latest restore log from security@kliper.dev.

Audit trail

Audit access for assessors

Firm administrators can export the complete audit trail (every field-level change, with actor, timestamp, IP address and user agent) for any date range, as CSV or JSON. It answers SOC 2 and QSA peer-review evidence requests. Audit export is part of the Firm plan.

Admin dashboard guide: audit export

Legal

Privacy & Terms

Privacy Policy: what we collect, how we use it, and your rights.

Terms of Service: the contractual basis for using the platform.

Changelog: every material change to subprocessors, security and this page is logged there.

Honest about what we don’t have yet

We are not SOC 2 certified yet, and we have no bug bounty program. The artifacts above (independent scanner grades, a public status page, a tested restore drill, an authoritative subprocessor list) are what we have built in their place while we are early. Questions? security@kliper.dev.