Security comes first.
RealMemory holds what your AI learns about the people it talks to. We protect it like it’s ours to lose: encrypted, isolated, under your control, and never used to train any model.
SOC 2 Type I and ISO 27001 in progress · Google Cloud and AWS · US and EU regions
Protected at every step.
From the moment a message reaches us to the moment you delete it.
Encrypted everywhere
Data is encrypted in transit with TLS 1.2 or higher, and at rest with AES-256.
Your memory stays yours
Each customer’s memory is kept separate from every other customer’s.
Personal data removed
We use Google Cloud’s Sensitive Data Protection service to find and redact personal information in messages, such as card numbers, before anything is stored.
No training on your data
Your memory is never used to train any AI model, ours or anyone else’s.
Delete a person everywhere
One request deletes a person’s data across all of their memories in near real time, with a full audit log.
View, correct, export
Your team can see, correct, delete and export any memory. It’s portable, so you’re never locked in.
Where your data lives.
Two major clouds, your choice of region, and access only for the people you choose.
Google Cloud and AWS
We run on both Google Cloud and Amazon Web Services for reliability. You choose where your data is stored: the United States or the European Union.
SSO and role-based access
Sign in through your company’s identity provider, and decide who on your team can see and change what.
AI model providers, listed
Some processing uses third-party AI model providers. Each one is listed as a subprocessor in our Trust Center.
Security is part of every job.
We set up our own company the way we’d want our vendors to: locked down first, then opened only where needed.
Strong sign-in everywhere
Every team account uses single sign-on and two-factor authentication. Company credentials live only in a managed password manager.
Zero-trust access
Internal tools are reachable only by verified team members through our zero-trust network, never from the open internet.
Checked and reviewed
New team members pass a background check before they start, and we review the security of our critical vendors regularly.
Everything in one place.
Our Trust Center, powered by Vanta, is coming soon. It will be where you can follow our certifications, download our security policies, review our subprocessors and request our DPA. Until then, email our security team and we’ll send what you need, including answers to your security questionnaire.
SOC 2 Type I
An independent audit of how our security controls are designed.
ISO 27001
The international standard for managing information security.
Policies, DPA and questionnaires
Security policies, our subprocessor list and our DPA, plus a place to send your security questionnaire.
Report a vulnerability.
If you believe you’ve found a security issue in RealMemory, email us at security@realmemory.ai. We read every report and will work with you to fix it. Please give us reasonable time to fix it before you disclose it to anyone else.
Give AI real experience.
Investing in AI infrastructure? Let’s talk →