Security you can build on
Encryption in transit and at rest, row-level rules enforced in the data layer, and a security scan before every publish. This page describes what is actually in place — nothing more.
Start buildingThe short version
Three commitments every app on GateBuild inherits from the platform itself.
Encryption
Traffic to and from GateBuild and your apps travels over TLS. Stored data — your app's records, files and backups — is encrypted at rest.
A scan before every publish
No version goes live unscanned. Every publish passes the platform's security scan and validation gates first; a failing check stops the publish until it is fixed.
Row-level security
Access rules are enforced in the data layer itself, not in the interface. A hidden button is not a permission; a rule the store enforces on every read and write is.
Your data, your rules
Three controls that belong to you, not to us.
Data residency
On Business and Enterprise plans you choose the region your workspace data is stored in, and it stays there. The choice is part of the workspace, not a support favour.
Business+Your data never trains our models
What you build, and what your users type into your apps, is used to run your apps — and nothing else. Training on customer data is off, for every plan, by design.
Deletion that reaches the data
Your apps get a user-data deletion flow, so when one of your end users asks to be erased, you can honour it — the records go, not just the account screen.
How we work
The practices below are part of how GateBuild is built and run today, each scoped to what it actually covers.
Secure development lifecycle
Security is a stage of building, not a review at the end: designs are checked for access mistakes before code is written, and changes are reviewed before they ship.
Validation before anything ships
Every build passes a seven-step validation gate and a twenty-point design gate before it can be published. A build that fails is not charged and does not go live half-checked.
Secrets never live in client code
API keys and credentials are stored as platform secrets and referenced by name. They are never embedded in the code your visitors' browsers download.
Payments stay with payment providers
Card payments are handled by PCI-DSS payment providers. Card numbers pass through the provider's checkout — GateBuild never stores them.
Scanning inside the publish gate
The publish gate scans dependencies and code patterns as part of every publish, so a known-bad library or an insecure pattern is caught at the door, not in production.
Snapshot before publish, rollback after
Every publish first takes a healthy snapshot of the live version. If the post-publish check fails, the platform rolls back to that snapshot automatically.
Access control on the platform itself
The same discipline applies to us: workspace roles decide who can open a builder account's apps, data and settings, and sensitive actions sit behind the roles that own them.
Incident response
A defined process for when something goes wrong: detect, contain, assess whose data is affected, notify the affected workspaces, and record what changed so it does not repeat.
Subprocessors
The companies that process data on GateBuild's behalf — our actual list, kept short and current.
Cloudflare
Hosting and DNS for the GateBuild site and published apps.
United States · global edge network
Zoho
Business email for the holding's mailboxes.
United States
Frontier-model API provider
AI generation: the rented model API the builder's brain runs on. The provider's name is published here when the launch key is activated.
Named at launch
Security for every app
The scan that runs before every publish checks five things. Every finding arrives with a severity, a plain-language explanation of what it means, and a fix the AI proposes for you to approve.
Vulnerable dependencies
Libraries your app relies on are checked against known vulnerabilities, so a component with a published flaw is flagged before your users ever load it.
Insecure code patterns
The generated code is scanned for patterns that leak data or trust the browser too much — the mistakes hand-written and generated code both make.
Exposed secrets
Keys and credentials are hunted wherever they should never be: in client code, in page text, in anything a visitor could download.
Missing login checks
Pages and actions that should sit behind a login are tested for exactly that — a screen that hides a button but serves the data anyway is a finding, not a pass.
Weak data rules
Row-level rules are reviewed for the classic gaps: a table anyone can read, a record anyone can edit. The store enforces the rules; the scan makes sure they exist.
Workspace security
How a team workspace protects itself, before enterprise controls enter the picture.
Roles
Owner, Editor and Viewer on every plan; Admin joins on the top tiers. A role decides what a member can open, change and publish — across every app in the workspace.
Sign-in options
Email and password or social sign-in, with email verification, anti-bot checks and optional two-factor authentication on paid plans.
SSO via OIDC
Connect your own identity provider — Entra ID, Okta, Google — so the workspace signs in where your company already does.
Business+Verified workspace domain
Verify your domain once with a DNS record, and access policies — including automatic SSO for your domain — can build on it.
IP allowlist
Restrict workspace access to your office and VPN addresses; requests from anywhere else are refused.
EnterpriseGovernance & monitoring
The controls larger organisations ask for, labelled with the tiers they belong to — never hidden behind a sales call.
SCIM provisioning
Members are created, updated and removed from your identity provider, so access follows your HR reality.
EnterprisePublishing controls
Decide who may publish, and to what visibility. A blocked publisher can request approval instead of hitting a wall.
Business+Connector management
Admins enable or disable connectors across the whole workspace, with the impact shown before the change is confirmed.
Business+Per-member credit limits
A monthly cap per member, so one person's heavy week cannot spend the workspace's whole pool. Owners and admins can raise it.
Business+Monitoring API
Usage, health and analytics flow into your own tools through scoped workspace API keys.
EnterpriseAudit logs
Sign-ins, publishing and governance changes are recorded, and can be streamed out through the Audit Logs API.
EnterpriseTrust Center
The evidence behind this page — policies, the scan methodology and the processor list — is kept in a gated Trust Center rather than posted publicly.
Security questions
Is GateBuild secure?
Security here is mechanics, not adjectives: TLS in transit and encryption at rest, row-level rules enforced by the data store itself, secrets kept out of client code, and a scan plus validation gates before every publish. The practices section above lists exactly what is in place today.
Which certifications do you hold?
None yet, and this page will not pretend otherwise. Independent certifications such as SOC 2 Type II and ISO 27001 are a launch-phase goal; when one is earned, it will appear here with its scope. Until then, judge us on the mechanics described above.
Where does my data live?
On the platform's hosting infrastructure, served through a global edge network. Business and Enterprise workspaces choose their storage region, and the subprocessor list above names every company that processes data for us.
How is data encrypted?
In transit, everything moves over TLS. At rest, stored records, files and backups are encrypted. Payment card data is never stored by GateBuild at all — it stays with the payment provider.
Can my company use its own SSO?
Yes, on Business plans and up: workspaces connect an OIDC identity provider such as Entra ID, Okta or Google, and Enterprise can enforce SSO for the whole workspace. Apps you build can offer SSO to their own users on the top tiers as well.
Is two-factor authentication available?
Yes — optional two-factor authentication (authenticator app or SMS) is available on paid plans for workspace sign-in, alongside email verification and anti-bot checks.
Can I stop the AI from changing parts of my app?
Yes. You can freeze files, pages and data definitions: frozen parts are never modified by the builder, and the publish scan treats an attempted change as a finding. It is how teams protect the pieces a human has already perfected.
Build on ground that holds
Scanned before it ships, ruled at the data layer, yours to export. Start with the free plan.
Start building