When a Digital Assistant answers your customers, it handles their questions, and sometimes their details. Buyers rightly ask how that data is protected. We’d rather give specific answers than security slogans, so here they are.

Each client in its own compartment

Every client runs in its own separate compartment, with its own software, database, storage and keys. Network rules block traffic from every other client. Your customers’ conversations are not stored alongside anyone else’s.

Messages can’t reach the wrong client

Every incoming message is routed only to the client that owns that channel. If a message arrives for a channel we don’t recognise, it is set aside and flagged to our team, never guessed.

Connected accounts encrypted with your key

When you connect accounts such as Gmail, Outlook, Meta or Zendesk, the login tokens are encrypted with a key unique to you before they’re stored. Only people allowed to manage that connection can use it.

Data stored in the UK

Your data is stored in the UK, in AWS London. Some services process data elsewhere, such as the AI model provider you choose. We’ll give you the list so you can decide what’s acceptable for your organisation.

Tested before it ships

Production only runs software that has already passed our staging environment. Every release is scanned for known vulnerabilities before it ships, and again every week.

Backed up, and checked

Every client’s data is backed up every night in two independent ways. The backup run checks itself and fails loudly if any client’s backup is missing or out of date.

Honest AI, approved boundaries

  • Your Digital Assistant says it’s an AI. You can show a notice before the first question, for example asking people not to share sensitive personal details.
  • It answers from your approved knowledge and only performs the tasks you switch on.
  • Anything that needs a person is handed over to your team.

Where we are on certification

We’re working towards SOC 2. We’ve mapped our controls against it and run automated checks on our own controls. We’ll tell you exactly where we are, and answer your security questionnaire line by line, in your first conversation with us.

For the public

If you’re talking to a Sum Vivas Digital Assistant on an organisation’s website, phone line, social channel or screen, you’re talking to an AI working for that organisation. It answers from information they have approved. Their privacy notice explains how your conversation is used.

Got a security questionnaire? Send it to us. We’ll answer it plainly.