Platform · Security & Operations
Prototyping is free. Operating is not.
Anyone can stand up an IoT prototype in an afternoon. Thinger.io is built for what comes after: corporate identity, device trust, backups, and someone accountable on the other end.
The problem
Production is a different product
An afternoon and an AI assistant will get you a working IoT demo. Nobody pays for that: they pay for the three answers between the demo and production.
- The demoAn afternoon
- ProductionWhat customers pay for
Pillar 01 · Identity
Corporate identity, per brand
Your customer's people sign in with the corporate account they already have, under the policy their security team already enforces. Pick a brand to see its policy.
- Password sign-in
- Disabled
- Identity provider
- Corporate directory, OpenID Connect
- Two-factor
- Required
- Password sign-in
- Enabled
- Identity provider
- Own accounts only
- Two-factor
- Required: authenticator code or security key
- Password sign-in
- Enabled
- Identity provider
- Google, picked by this brand
- Passkeys
- Offered on the sign-in screen
Identity on the platform In Brand A's policy
Sign in with the company account
Federation over OpenID Connect to Google, Microsoft Entra ID, Okta, Auth0, Keycloak or any compliant provider.
OpenID Connect
Accounts on your terms
Provision users just in time under a domain allow-list, or run match-only so only accounts created in advance may sign in.
JIT provisioning · Match-only
Two-factor beyond passwords
Time-based one-time codes and WebAuthn security keys: the second factors corporate security teams already require.
TOTP · WebAuthn
Passkeys
Passwordless sign-in, built into the platform sign-in itself.
Passkeys
Access that stops where you say Every brand
Roles combine allow and deny rules down to single resources and actions, projects scope what each team sees, and API tokens carry least-privilege scopes you can revoke.
Allow / deny · Projects · Scoped tokens
Brands live on a white label instance: each of your OEM customers gets its own domain, branding and corporate directory, on one instance you operate.
Pillar 02 · Device trust
Every device proves what it is
A fleet is only as trustworthy as its least trustworthy device.
The platform treats each one as its own principal: its own secret, its own encrypted channel, and, when your deployment demands it, its own certificate.
A secret per device
Own credentials per device, stored as salted hashes. A leak affects one device and revoking it touches nothing else.
PBKDF2 · SALTED HASHES · PER DEVICE
TLS on every transport
HTTPS, secure MQTT and the native IOTMP protocol all run encrypted.
HTTPS · SECURE MQTT · ENCRYPTED IOTMP
Mutual TLS with X.509
Devices can prove their identity with a client certificate, issued by your own CA if you prefer.
X.509 · CLIENT CERTIFICATES · YOUR OWN CA
Certificates that renew themselves
Let’s Encrypt certificates are issued and renewed automatically, on every domain of a white label.
ACME · AUTOMATIC RENEWAL · MULTI-DOMAIN
OPERATED, NOT JUST DEPLOYED
Pillar 03 · Continuity
Continuity of operations
Whether an incident becomes an anecdote or a crisis is decided long before it happens.
Backup and restore, platform level
One consistent snapshot covers the device registry, time-series data and platform assets, and ships encrypted to object storage. Offered as a managed service on cloud instances, activated per instance.
- Consistent snapshot
- Encrypted
- Object storage
The instance watches itself
CPU, memory, disk, network and pending OS security updates, monitored natively in the platform and through a lightweight agent on the host. You see the machine, not just the software.
- CPU
- Memory
- Disk
- Network
- OS updates
Environments rebuilt from code
Every managed instance is deployed as infrastructure-as-code, so an environment can be rebuilt from its definition. That is the backbone of disaster recovery: a repeatable procedure, not heroics.
- Infrastructure as code
- Reproducible
Same software everywhere
The identical platform runs in the community cloud, private cloud, on-premise and enterprise deployments. Moving between them changes where it runs, not how it is secured.
- Cloud
- Private cloud
- On-premise
- Enterprise
Certification
Security is a process, not a feature list
Thinger.io is ISO 27001 certified as a company: the certification covers the whole organization and the processes behind the platform, not just the software.
Plan Decide what to protect, and how.
- Risk assessment
- Security policies and controls
Do Run the controls, day to day.
- Dependency patching
- Key and certificate rotation
- Least-privilege access
Check Prove they still work.
- Access reviews
- Backup restore tests
- Internal and external audits
Act Fix what the checks find.
- Incident response
- Corrective actions
- Lessons fed into the next plan
Without the loop
- Certificates expire
- Dependencies age
- People change roles
- Backups go untested
A certified management system is the discipline that keeps a security page true after it is published.
For your security review
Everything your compliance team will ask for
Security and procurement reviews come with a checklist. These are the items we are asked for most, and how you get each one.
security@thinger.ioQuestions, documents and vulnerability reports- On request
ISO/IEC 27001 certificate
Issued to the whole organization, not only the platform. We share it and its scope with your team.
- On request
Data processing agreement
A GDPR agreement for the personal data your users and devices send to the platform.
- Your choice
Data location
Our cloud, a dedicated instance in the AWS region you choose, or your own servers.
- We answer
Security questionnaires
Send us the one your company uses and we answer it.
- On request
Penetration testing
Tests against your deployment when your policy requires them.
Good to know
Frequently asked questions
What security teams ask before production, answered plainly.
Do you support SAML?
No. Federation is OpenID Connect only. In practice, every major corporate identity provider speaks OIDC: Microsoft Entra ID, Google Workspace, Okta, Auth0 and Keycloak all federate with the platform today. If your identity provider can only speak SAML, we would rather tell you before the contract than after.
Can each of our customers bring their own identity provider?
Yes. Identity providers and authentication policy are configured per white label brand. Each brand on your instance can federate against its own corporate directory, disable password login to enforce SSO, require two-factor for its users, and choose which providers appear on its own sign-in screen.
Where does our data live?
On your instance, in the region you choose when it is deployed. Managed cloud instances are dedicated servers, and EU regions are available for deployments that need European data residency. Private cloud deployments keep the data inside your own cloud account, and on-premise keeps it inside your building.
Can we run it fully inside our own network?
Yes. On-premise deployments run the identical platform on your own infrastructure, inside your own network perimeter. The security model does not change with the deployment model, so what you evaluated in the cloud is what you run at home.
Is platform backup enabled by default?
No, it is activated per instance. Platform-level backup is offered as a managed service on cloud instances: when active, it takes one consistent snapshot covering the device registry, time-series data and platform assets, and ships it encrypted to object storage. If you want it on your instance, ask us and we will set it up.
Ready for the day it becomes production
Tell us about your users, your devices and your deployment, and we will map them to the features on this page. Running a security review? We will answer your questionnaire and share our certificate.