Security and compliance

How Zembula protects your data, your subscribers and every email we personalize.

Two trust badges on a dark ground. SOC 2 Type II: audited security controls. GDPR and CCPA: data privacy, EU and California.

Brands trust Zembula with data about their subscribers every time an email opens. We protect that data with one security program that covers our people, our infrastructure and our platform. We collect only the data that personalization needs, and we encrypt it in transit and at rest. An independent firm audits our controls against SOC 2 Type II every year. An independent security firm tests our platform by hand.

This page summarizes how our program works. Security and procurement teams can request our SOC 2 report and other security documents through the Zembula Trust Center. Live availability is public on our status page.

Compliance and independent testing

Independent firms check our controls and test our defenses. You can review their reports under NDA.

SOC 2 Type II

Zembula completes an annual SOC 2 Type II audit by an independent audit firm. Request our most recent report, under NDA, through the Trust Center.

Penetration testing

An independent security firm tests our web application, API and cloud configuration by hand, the way a real attacker would.

Privacy regulations

Our privacy program and our Data Processing Agreement (DPA) address GDPR, UK GDPR, the CCPA and other US state privacy laws. They also address Canada’s PIPEDA and Quebec’s Law 25.

Continuous monitoring

Zembula uses Vanta to test our cloud configuration, company computers, user access and policies continuously. Vanta flags each failed test for us to fix.

Security policies

Our information security policies cover access control, cryptography, data management, incident response, business continuity, secure development, vendor management and the use of AI. We review and approve them at least once a year. Every team member must accept them.

Privacy and data protection

We design the platform to need as little personal data as possible, and we protect what it holds.

We collect only what personalization needs

Zembula identifies people by pseudonymous identifiers, such as an irreversible hash of an email address. You choose which personalization fields to send us, and we never copy full subscriber profiles. Our event data describes products and actions, not people. We never collect payment card data.

Rendered images are never stored

Zembula renders each personalized image fresh when the email opens. We never store rendered images, and we never reuse one recipient’s image for another.

Where your data lives

We store customer data in the United States, on Amazon Web Services (AWS) and Heroku (Salesforce). Personalized images render at the AWS edge location nearest each recipient.

Encryption

All connections to Zembula use HTTPS with TLS 1.2 or higher. We encrypt customer data at rest with AES-256, including backups. Credentials that you store in Zembula, such as API keys for your data sources, get a second layer of encryption with keys held in AWS Key Management Service. After you save a credential, nobody can read it back, including Zembula staff.

Retention and deletion

Behavioral data expires automatically on a set schedule. When you ask, or when your contract ends, we delete your personal data within the period that our DPA sets. We also delete an individual subscriber’s data on request, and we keep a record that proves the deletion.

Data Processing Agreement and sub-processors

Zembula acts as a processor of your data. Our DPA includes the EU Standard Contractual Clauses and the UK Addendum for transfers of EU and UK personal data. Ask your Zembula contact for a copy.

AI

The AI assistant in the Zembula platform runs on Anthropic’s Claude models. Anthropic’s commercial terms do not let it train models on our customers’ content. Our AI Governance Policy sets which AI tools our team may use with customer data.

How we secure the platform

One security program covers our people, our infrastructure and the way we build software.

People

Every new team member must pass a background check before they get access to production systems or customer data. Each person completes security awareness training when they join and every year after that. Company computers run software that checks disk encryption, screen lock and other security settings.

Access control

Access to production follows least privilege. Privileged access requires multi-factor authentication, and we review user access every quarter. Zembula staff sign in through single sign-on, and we remove access when someone leaves.

Infrastructure and network

We run no data centers of our own. Our platform runs on AWS and Heroku, whose data centers hold SOC 2 and ISO 27001 certifications. Our core production services run in a private network across three AWS availability zones. A web application firewall protects the main application, and internal services cannot be reached from the internet.

Monitoring and detection

Amazon GuardDuty watches our AWS environment for threats, and every medium, high or critical finding pages our on-call engineer. We log all traffic in our production network. Alarms also page on-call staff when availability or data completeness drops.

Vulnerability management

Automated scans check our code dependencies and our cloud configuration for known vulnerabilities. We fix each vulnerability within a set deadline based on its severity.

Secure development

Every code change goes through version control, protected branches and a reviewed merge. Production deploys run automatically, and each deploy ends with a health check against production.

Vendor management

We review the security of each vendor before we share customer data with them. Every vendor stays in a monitored inventory with its risk rating.

Reliability and incident response

Our standard service level agreement (SLA) commits to 99.95% availability for image rendering and 99.5% for the Zembula platform, measured each calendar month.

Live status

See live availability and 90 days of history at status.zembula.com. The status page and our SLA reports use the same data, so the two always agree.

Built to fail safe

Every personalized image has a fallback. If an image cannot render, the email shows the fallback, so recipients do not see a broken image. A Zembula outage cannot block or delay your sends.

Backups and disaster recovery

Our primary databases have continuous backups with point-in-time recovery, and all backups are encrypted. Every year we test our restores in a disaster recovery exercise. We also run tabletop exercises twice a year.

Incident response

Our Incident Response Plan defines severity levels, escalation paths and how we keep customers informed. If a breach affects your personal data, we notify you within 48 hours of confirming it. We then send updates as we learn more.

Security controls for your team

Your administrators decide who gets in and what each person can change. Some features depend on your plan.

  • Single sign-on: SAML 2.0 and OpenID Connect, with Okta, Microsoft Entra ID, Ping Identity or Google Workspace. You can make SSO the only way to sign in.
  • User provisioning: SCIM 2.0 creates, updates and deactivates users from Okta or Microsoft Entra ID.
  • Two-factor authentication: require it for every user on your Zembula account.
  • Roles: only Admins can manage data sources, credentials, SSO and integrations.
  • Audit log: every change records who made it, when, and the values before and after.
  • Alerts: send audit events to email, or to a signed webhook that Splunk can ingest.
  • Content locks: Admins can lock published content so that nobody can change it.
  • Signed image links: each image link carries a signature, so nobody can change what an image shows by editing its link.
  • Custom domains: serve content from your own subdomain, with certificates that renew automatically.

Contact us

Performance marketing dials, inside email. For the first time.

See how we do it by booking a demo.