Trust · Security
How we handle your data, in writing.
Where your data sits, who can reach it, how long it stays, and who answers for what the AI does. We hold no security certifications, and this page says so before it says anything else.
In the contract, before the build
Every line here is a clause you can hold us to.
In short
7 Minds Systems holds no security certifications. What we sign up to is written into the contract for your build. Your enquiries stay in your own CRM and calendar. Each AI employee gets only the access its job needs. Your documents are read at the moment a question is asked, and no model is retrained on them.
What this page is
Every promise on this page is a clause we sign.
A security page usually opens with a wall of badges. This one cannot, because 7 Minds Systems holds none. What we have is a set of commitments written into your contract and a build spec that says, role by role, what each AI employee may touch.
Your enquiries and client records stay where they already live, in the CRM and calendar your team logs into. We build the workforce around those systems, and any provider we introduce is named for you in writing before the build starts.
Each AI employee is scoped the way a job is scoped. It has a defined remit, the access that remit needs, and a named person it escalates to. Who reviews its output, and when, is set out in how we work.
The legal detail lives in the privacy policy, which sets out the categories of processor we use and the safeguards on any transfer outside the UK. The named sub-processor list is available on request. If your compliance team has a question this page does not answer, ask it on the first call and the answer goes into the spec.
How it is actually run
Four controls, set before a single enquiry arrives.
These are configuration choices, made for your build and recorded in the spec you sign. None of them depends on a badge, and none of them changes without your approval.
- 01
Your systems hold your records
Enquiries flow into the CRM and calendar your team already logs into, so client data never pools in a shared 7 Minds database. You keep the master copy, and the delete button stays yours.
- 02
Least privilege, role by role
Each AI employee is given the records and tools its own job needs and nothing past them. The map of who reaches what is written into the build spec and reviewed with you before go-live.
- 03
Retrieval, so your documents stay documents
Your handbooks and price lists are read at the moment a question is asked. No model is retrained on them, and nothing an AI employee touches is used to train a public model.
- 04
A person on the decisions that carry risk
You set the line between what the workforce may do alone and what it must hand to a human. Anything sensitive, complex or off-script goes to a named person with the context attached.
Asked on every first call
What happens if something goes wrong, or you want us out.
How do we revoke your access, and how fast does it take effect?
From your own admin panel, and immediately. The connections into your CRM, calendar and inbox are accounts you create and own, scoped to the permissions you set, so switching one off is your action and needs no request to us. Nothing we run keeps a private copy of your records to fall back on, because those records were in your systems the whole time. Every provider account opened for your build is named in writing before the build starts, so the list of things to switch off is one you already hold.
What happens if there is a security incident on our build?
You hear from us, and you hear the unflattering version. We do not run a 24-hour security operations centre and will not claim one to sound larger than we are. What we do commit to is the notification duty written into the data processing agreement. Where we act as your processor we tell you without undue delay, so your own reporting clock to the ICO starts with the facts we actually have. You get what we know at the time, and a written account of what changed once it was closed.
What do we keep if we stop working with you?
Your records, because they were never anywhere else. Enquiries, bookings and client notes sit in the CRM and calendar you own, so ending the engagement does not move them. You keep the build spec as well, which documents what each AI employee was allowed to reach and who approved its output. What stops is the workforce we run. The connections into your systems are closed, and anything we hold as your processor is deleted or returned on the terms set in the data processing agreement. We hold nothing back to make leaving harder.
Proof
Where the system runs today: our own group companies.
- Eta Medical Group
- AH Group Holdings
- Eta Investment Group
- AH Strategic Holding
- Eta Properties
Named with permission. Every firm here is a fellow member of the private group of operating companies that 7 Minds Systems belongs to, spanning healthcare, property and professional services. Verified results and testimonials are published here as each engagement signs them off, never before.
Bring your compliance questions to the call.
Book a thirty-minute call and we will answer them against your own systems and your own regulator, then put the answers into the build spec.
Thirty minutes, no pitch. Bring your data protection lead and we will take the questions as they come.
Questions
The data questions, answered.
Is our data used to train an AI model?
No. No model is retrained on your material, and nothing an AI employee touches is used to train a public model. Your handbooks, price lists and the replies you already give clients are held as a corpus the AI reads from, and the passages bearing on a question are put in front of the model at the moment it is asked. Edit a document and the next reply reflects the edit. Which sources are connected is agreed in writing before anything is wired up.
Which systems and providers hold our data?
Your enquiry and client records live in the systems you already run, so a booking lands in your calendar and a lead lands in your CRM. Any platform we add around them is named for you in writing before the build starts. Our privacy policy sets out the categories of processor we use, and the named sub-processor list is available on request. Where your regulator or your clients expect UK or EU data residency, we configure your build to process and store enquiry data in that region.
How is access controlled role by role?
By least privilege, applied per role. Each AI employee is given the specific records and tools its job needs and nothing beyond them, so the AI employee that books appointments is not handed your finance folder. Your own team keeps the permissions your CRM and calendar already enforce. The map of which role reaches which system is written into the build spec and reviewed with you before go-live, and it changes only when you approve a change.
How do we get our data deleted?
You ask, and we delete. Deletion and retention are both fixed in the data processing agreement we sign at the start of the engagement. You tell us what to remove. We remove it from the systems we run on your behalf and confirm in writing when it is done. Where we act as your processor, retention follows your instructions. Records held in your own CRM and calendar stay under your control throughout, so you can delete those yourself at any time.
Who at 7 Minds Systems supervises the AI?
Our team does, day to day, and you hold the sign-off. We monitor the conversations an AI employee has, correct the edge cases, and tune the voice as your business changes. You set the boundary between what it may do on its own and what it must hand over, and one named person on your side owns each escalation. In regulated work, every regulated step is approved by a human before it goes out.
What security certifications do you hold?
None. We hold no SOC 2 report, no ISO 27001 certificate, no HIPAA attestation and no Cyber Essentials badge, and we will not imply otherwise to win the work. What we put in writing is checkable. A data processing agreement covers lawful basis, retention and deletion. A named sub-processor list is available on request. A build spec states exactly what each AI employee may access and who approves its output before it goes out. Everything past that is a setting on your build, agreed with you and written into the contract before we start. If your build calls for UK or EU data residency, we configure it that way and name the providers. If it calls for provider accounts configured so that your prompts and outputs are not retained, we check what your chosen provider supports and configure that for your build too. Neither is switched on by default, and where a provider cannot support what you need, you hear that before you sign.
The full legal detail is in the privacy policy, and who signs off what is set out in how we work.
Book a strategy call
Thirty minutes to put your data questions to us.
Thirty minutes to scope this role properly: what it would answer, what it would escalate, and what it would cost to run.
Know the scope and the number before you start.
The build is scoped and priced in writing before it begins, with the exclusions named, and the AI workforce is built and trained on your services and voice before it answers anyone. You commit to a defined piece of work, not to a promise.
Book a thirty-minute call
Pick a time straight from the diary. If you would rather write first, email us and a person replies, usually the same working day.