Data processing

A LUNTA-built website talks to your customers, so the data question deserves a straight answer before the first call, not a clause after signature. This page states how every engagement is set up. Each commitment here is one you can hold us to in the data-processing agreement.

Where the data lives

Visitor conversations, leads, and bookings live in your systems, under your controls — your CRM, your calendar, your mailbox. We hold scoped operational access to run the site, and that scope and its expiry are written down, not assumed.

What the agents may say

The agents answer only from material you approved. They never invent a price, promise, or policy, and they hand off to a human the moment judgment is needed. Every conversation is logged, and the log is exportable and yours.

What we never do

• No model training on your customer data. Your customers’ conversations improve your site’s approved content through the tuning cycle you approve — they are never used to train models.
• No resale, no sharing, no enrichment. Lead data exists to serve your customer and your pipeline, nothing else.
• No silent tooling. The AI services used on your site are declared to you, and changes go through you first.

The paperwork

Every engagement that touches personal data runs under a GDPR data-processing agreement signed before launch — one you can hand to your lawyer. It names the processing purposes, the sub-processors (hosting and AI providers), and the exit terms: when you leave, your data leaves with you, and leaving is a download, not a negotiation.

Questions

Ask us anything about this before an engagement: hello@lunta.ai. If your security or legal team has a questionnaire, we answer it as part of the blueprint week, not as a change request.