Skip to content

Enterprise security camera systems without the enterprise contract

This page is for the person who signs the invoice and owns the risk. It answers the questionnaire before you send it, including the two questions where the honest answer is not yet.
  • No biometrics, by design
  • SSO, roles and audit logs
  • DPA and sub-processor list
01 The sentence you take to counsel

No biometrics. Not off by default: absent from the design.

Motiondetection classifies object categories: person, vehicle, animal, package, other motion. It does not build face templates. It does not match faces. It does not do gait analysis, and it does not re-identify a person across cameras or across days. No biometric identifier is computed, so none is stored, so none can be requested, subpoenaed or breached.

In Illinois, Texas and Washington, biometric collection is the trigger for the strictest consent and retention statutes in the country, and it is the reason a lot of camera projects die in legal review. A system that never computes one is a materially different purchase, and this is a product decision rather than a configuration you can accidentally reverse.

The mechanism behind that claim, including exactly what the model is sent, is on artificial intelligence video surveillance.

Facial recognition not present
Face templates or embeddings stored none
Gait or body shape identification not present
Re-identification across cameras not present
Licence plate recognition not present
Object category classification yes, this is the product
Confidence value on every event yes
02 The four things the questionnaire asks

Access, auditability, data handling, procurement. In that order, every time.

Access control

  • Single sign on over SAML 2.0 and OIDC, from the Multi-Site tier.
  • Roles: Viewer and Operator from Operate, Admin and custom roles from Multi-Site.
  • Per site scoping, so a contractor sees one yard and not the estate.
  • Session control and forced sign out on role change.

Auditability

  • Who viewed which clip, who changed which zone, who exported what, with a timestamp and an actor.
  • 30 days of audit log on Operate, one year on Multi-Site, custom on Enterprise.
  • Export of the audit log for your own retention system.
  • Alert delivery is logged too, so a missed page is answerable after the fact.

Data handling

  • Encrypted in transit and at rest.
  • A stated retention window per tier, with configurable deletion inside it.
  • A data processing agreement, available from Operate.
  • A published sub-processor list, with notice before it changes.
  • The video itself stays where your cameras are. Only event crops are sent for classification.

Procurement

  • Annual invoicing and purchase orders on Multi-Site and Enterprise.
  • W-9 on request.
  • Security questionnaires answered rather than deflected, by email.
  • A custom MSA and an on-premise connector are Enterprise terms.

Which of these arrives at which tier is set out line by line on the video analytics software price page.

03 What we are not claiming

Three claims a vendor would normally make here, and why this page does not make them.

No compliance certification is claimed

We are not going to put a badge on this page for an audit that has not happened. If your process requires a completed SOC 2 report or an ISO 27001 certificate before signature, say so by email and we will tell you honestly where that stands rather than implying it is done.

No customer names, logos or references

We do not publish who runs it, so there is no trusted by strip anywhere on this site and no case study with the name removed. A logo wall we do not have is not a design gap, it is a lie we chose not to tell.

No accuracy percentage

Any single number here would be measured on footage that is nothing like yours. Instead every event carries its own confidence value, the conditions where the model gets worse are named on the how it works page, and the demo runs on your own clip so you can measure it on the scene you care about.

04 The arithmetic, in your units

The failure you are buying against is the muted alert channel, not the missing camera.

At 15 cameras, an enterprise platform positioned at about $100 per camera is roughly $18,000 a year, most of it paying again for hardware you already own. Operate is $1,428 a year. Both numbers are checkable: ours is published on this site, and theirs is their own public positioning rather than a quote we have seen.

The part that does not appear on either invoice: your operator muted the alert channel in week two, because it fired forty times a night at rain. Every camera on the site is still recording and nobody is watching. That is the failure mode the purchase is actually against, and it is a software problem with a software price.

Run the numbers with your own camera count on the pricing page, and if the answer needs a purchase order, that is a Multi-Site or Enterprise term.

What to put in the approval note

  • No new hardware, so no capital request and no installer.
  • No biometric data collected, so no consent regime under the state statutes.
  • SSO, roles and an audit log at a named tier, not as a future promise.
  • A DPA and a published sub-processor list.
  • An annual price you can read on a public page, in writing, before any call.
  • A cancellation that is a monthly plan ending, not a multi-year term.
05 Asked in security review

Four questions that come up in every review.

01

Where is the data processed?

Detection runs where the footage is: in the browser in the demo, and next to your cameras on a live site. Classification runs on small still crops sent to a model provider. The sub-processor list names who that is, and it is published rather than described.

02

Can we run it entirely on our own network?

An on-premise connector is an Enterprise term. On every other tier the detection is local and the classification call leaves the network, which is a distinction worth being precise about in a review rather than glossing.

03

What happens to our data if we cancel?

Event history is deleted on the retention schedule of your tier, and you can trigger deletion earlier. Cancelling stops processing; it does not put your footage anywhere new, because your footage was never ours.

04

Who do we contact about a vulnerability?

The same address as everything else on this site: support@motiondetection.com. There is no contact form anywhere here, on purpose. Mail sent to that address is read by the people who build this.

Send the questionnaire to a person, not a portal.

support@motiondetection.com is read by the people who build this. Or confirm an email address here, pick the tier by camera count and the console opens.