QRMYour data, protected and in your control
QRM is built for organizations that take security seriously. Our own people test it the way an adversary would, named people answer for it, and you decide where your data lives.
- Built to SOC 2 criteria
- Built to ISO 27001
- Reviewed every month
- Deploy in any region
- Never trains on your data
Last reviewed November 2, 2026
On this page
Our commitment
Every decision we make about QRM starts with three commitments.
Privacy
Your data is yours. We do not sell it, we never use it to train AI models, and you choose where it lives.
Security
Defense in depth at every layer, reviewed every month by our own staff and tested by our own red team, with every confirmed finding tracked to a fix.
Trust
Every new hire passes a pre-employment background check, and most of our staff hold active SECRET or TOP SECRET clearances. Our staff use MFA on email, source code and every cloud console, and complete security training every year.
Deploy anywhere
Your data lives where you decide
Run QRM in your own cloud account on Amazon Web Services, Google Cloud, Microsoft Azure or Oracle Cloud, in any region you choose. Or run it in your own data center. Your data stays in the location you select.
Your cloud
A dedicated QRM deployment in your own cloud account, in the region you choose.
Your data center
QRM on your own servers, behind your own physical and network controls.
QRM cloud
Fully managed workspaces at yourteam.qais.app, ready in minutes.
- Amazon Web Services
- Google Cloud
- Microsoft Azure
- Oracle Cloud
Every Enterprise deployment is dedicated to a single customer.
Security
- Sign-in with single sign-on and MFAVerified
- Sign-in refused at the second factorDenied
- Role changed by an administratorRecorded
- QaiS prepared an email to a customerAwaiting approval
- A person approved the email, then it was sentApproved
Encryption
Data is encrypted in transit with TLS 1.2+ and sensitive data is encrypted at rest with AES-256.
Enterprise access controls
Single sign-on, multi-factor authentication, role-based access and an append-only audit log of sign-ins, failed sign-ins, administrator changes and agent actions, managed by your administrators.
Tenant isolation
Each workspace's data is kept separate from every other customer's, and every request is scoped to its own workspace.
Tested backups
Encrypted, off-site backups with point-in-time recovery, and full restores run by our staff to prove they work.
Red team
Our own people attack QRM, and named people answer for what they find.
Our red team is our own staff. They review the code and configuration every month and attack QaiS the way an adversary would: leaking data, overriding its instructions, turning it against the people it works for. Blane Steckline, our security point of contact, owns the security program and tracks every confirmed weakness to a fix.
Monthly security reviews
Our own staff review the code and configuration in depth every month, and every finding is tracked to a fix.
Adversarial testing of QaiS
Our staff attack QaiS on the live chat surface with known prompt-injection and jailbreak techniques, from hidden instructions in a file to slow multi-turn escalation. Every confirmed weakness is tracked to a fix.
Recovery and response drills
Our staff run full restores from our off-site backups to prove they work, and rehearse the incident response plan with a tabletop exercise at least once a year. Customers are notified within the contracted time.
Every change scanned
Each code change is scanned for exposed credentials before release, and dependencies are checked every day for known vulnerabilities. Our staff fix critical and high findings within 30 days.
- Every monthCode and configuration reviewAn in-depth review by our own staff, every finding tracked to a fix.
- Oct 5, 2026Incident response tabletop exerciseOur staff rehearse the incident response plan at least once a year.
- Oct 2026Full restore from off-site backupsRun by our staff to prove the backups work.
- Sep 2026Adversarial run against QaiSKnown prompt-injection and jailbreak attacks on the live chat surface, with confirmed weaknesses tracked to a fix.
- Every yearSecurity trainingFor every person, at hire and every year, tracked.
Owned by Blane Steckline, security point of contact.
AI and your data
We don't train the model. We teach the harness.
The harness is everything around the model: memory, preferences, skills and the rules your people set. It learns how your workspace works and keeps that learning in your workspace, not in anyone's model.
- Your people work Corrections, ratings and "remember this"
- The harness keeps the lesson Notes, preferences and skills, stored in your workspace
- A nightly review The day's corrections become notes, with an Undo card each morning
-
QaiS works your way
Every answer recalls what your people taught it
QaiSNever trains on your data
Your data is never used to train AI models.
Agents follow your rules
AI agents act only with the permissions your people have. Every action is recorded, and sending, publishing or spending waits for a person's approval.
You stay in control
AI stays inside your workspace, and you decide which AI features are turned on.
Compliance
QRM's security program is built to the SOC 2 Trust Services Criteria and ISO/IEC 27001, and the platform runs on infrastructure from providers certified to both. Enterprise customers can run QRM inside their own cloud account. Send us your security questionnaire and we complete it.
Infrastructure certifications
A current list of sub-processors is available to customers on request, under NDA.
Security contact
Questions about security, questionnaires or documentation? Email security@qais.app.
- Blane StecklineDirector of SoftwareSecurity point of contactAnswers security questions, questionnaires and vulnerability reports.
- Nick EllisDirector of Engineering
- Qais AlkurdiChief Executive Officer
Responsible disclosure
Found a vulnerability? Our security contacts want to hear from you.
Email security@qais.app with "Security" in the subject and include the steps to reproduce.
Please
- Test only against a workspace you own or are allowed to test.
- Never access, change or delete other people's data.
- No denial-of-service, social engineering or physical attacks.
- Give us reasonable time to fix the issue before sharing it.
In return
- We confirm your report and keep you updated.
- We credit you when the fix ships, if you wish.
- We do not take legal action against good-faith research that follows these rules.
Common questions
Can we choose where our data is stored?
Yes. Deploy QRM in your own cloud account, in any region you choose, or in your own data center.
What is QRM's compliance status?
QRM's security program is built to the SOC 2 Trust Services Criteria and ISO/IEC 27001, and the platform runs on infrastructure from providers certified to both. Enterprise customers can run QRM inside their own cloud account. Send us your security questionnaire and we complete it.
Is our data used to train AI?
No. Your data is never used to train AI models.
How do you test QRM's security?
Our own staff review the code and configuration every month, scan every change before release and attack QaiS with known prompt-injection and jailbreak techniques. Every confirmed finding is tracked to a fix, and we welcome reports from outside security researchers.
What happens to our data if we leave?
Administrators can export workspace data, and account owners can request deletion.
Can you complete our security questionnaire?
Yes. Send it to security@qais.app.