Security

Vestly talks to your customers on your behalf. So we keep their data the way you would want your own kept: in Europe, encrypted and only as long as needed.

Your data stays in Europe

• Servers and the database are in Hetzner's data centre in Helsinki (Finland).
• Encrypted backups are in Backblaze's EU data centre in Amsterdam (Netherlands).
• To write an answer, only the question, the start of the conversation and the parts of your website that relate to it go to the AI provider, never the whole database. AI providers may process it in the United States, under the European Commission's standard contractual clauses and/or the EU–US Data Privacy Framework. Every subprocessor is in the list of subprocessors.

AI and your data

We do not use your data or your customers' data to train AI models, and we do not sell it. Under their terms, OpenAI and Anthropic do not train their models on data sent through their APIs. The AI answers only from your website and knowledge base; when the answer is not there, it says so or brings in a person.

Encryption

• All traffic uses HTTPS (TLS): the chat window, the dashboard and the API.
• Passwords are hashed. Integration keys and access tokens (Google Calendar, Shopify, Meta, CRMs) are encrypted in the database (AES-256).
• We do not store visitors' IP addresses, only a salted hash to stop abuse.

Access

• The dashboard has four roles: owner, admin, agent (answers chats and works leads) and viewer.
• Two-step sign-in with an authenticator app, with recovery codes.
• On Enterprise, your team can sign in only with your company's Google Workspace accounts: when someone leaves and their Google account is closed, their access ends too.
• The server accepts SSH keys only; password and root logins are off, and the firewall opens only the ports needed.

Backups and reliability

• The database is backed up every night, encrypted (age, X25519) and moved to a data centre in another country. Daily backups are kept 35 days, monthly ones 13 months. Restoring has been tested.
• External checks test every 5 minutes that the service answers; we are alerted at once when it does not.
• When an AI provider is slow or down, a second one answers the chat.

Retention and deletion

Conversations are deleted after 90 days, photos visitors send after 30 days and leads after 24 months; sooner if you choose. Any conversation or lead can be deleted at once in the dashboard. When an account closes, we delete its data within 30 days. The WordPress plugin adds Vestly's data to WordPress's personal data export and erase tools.

Transparency

• The activity log shows who changed settings, took over a chat, exported leads or created an API token.
• When our support looks into your account with your permission, that is logged too.
• The data processing agreement is ready: the account owner accepts it in the dashboard and gets a signed PDF.

For clinics and others with sensitive customer data

• Set the chat to ask only for a name and contact details, not health information. You choose the extra fields of the contact form.
• Turn off photos in the chatbot's privacy settings if a picture could show health details.
• Shorten how long conversations are kept if needed (privacy settings, in days).
• The chat gives no medical advice: it answers about prices, services, times and location, and refers to a doctor.
• When a visitor writes about their health anyway, you can delete the conversation at once in the dashboard.

Enterprise

An Enterprise agreement sets the availability level (SLA), response times for incidents and service credits, the data location, and we fill in your security questionnaire if needed. SAML sign-in is available on request.

Reporting a security issue

If you find a vulnerability, write to hello@rixio.ee. We answer within one working day and ask you not to publish it before it is fixed.