MetoCRM

Medical Tourism AI Agent

Through MetoCRM's LLM gateway an AI assistant can read your CRM, pull reports and act on your behalf. Personal data never passes through it.

What it is

A medical tourism AI agent is a language-model assistant connected to your MetoCRM data over a standard protocol. It does not take screenshots or imitate the interface; it reads patients, conversations, quotes and reports directly through the API and performs the actions it is permitted to.

In practice: someone on your team writes "list patients who were sent a quote this week but haven't been touched for three days" and gets the list. Or "break down last month's leads by channel" and reads the report. No intermediate step, nobody logging in to build a filter.

What it can do

  • Read patients and conversations — List and filter patients, open a patient record, appointments, conversation history and messages.
  • Send messages — Reply into a conversation; read the template list and use the right one.
  • Track quotes — List quotes and quote templates, read the status of a quote.
  • Consultations — Open a consultation, send messages and update its status.
  • Pull reports — Sales, leads, calls, CRM messages and advertising reports — broken down by channel and period.
  • Update records — Update permitted fields on patient and user records.
  • Reference lists — Statuses, tags, languages, branches, advertising channels, user groups, currencies.

Personal data does not pass through

This is the gateway's design, not a setting added afterwards. The JSON it returns never contains:

  • Patient and user email addresses or phone numbers
  • Conversation identity details
  • Contact fields on lead records

So the assistant can see a patient's treatment request, status and conversation flow — but not what it would need to reach them. When it sends a message it does so through the CRM's own channel, not by learning a number.

Health data is a special category under both GDPR and KVKK. When evaluating any AI integration the first question should not be "what can it do" but "what leaves the building".

What it is actually used for

  • Morning briefing — "Yesterday's lead count, unowned records, and follow-ups due today" — the assistant writes the stand-up agenda.
  • Finding files that went quiet — Listing records where a quote was sent with no reply, or whose next step is overdue. Done by hand, nobody does this every day.
  • Channel performance questions — "How does Instagram's enquiry-to-quote rate compare with paid lead forms?" — read from the report endpoints and interpreted.
  • Conversation summaries — A summary of a long patient thread and a suggested next step; saves minutes at every handover.
  • Consultant performance — Open records, quotes and closes per person — comparative interpretation rather than raw numbers.

How it connects

The connection uses a standard protocol supported by language-model clients. A key is issued for your own account, and the assistant can reach only that tenant's data. Revoking the key closes access immediately.

For setup and permissions, talk to our sales team; the gateway is closed by default on every account and opened on request.

Frequently asked questions

Will the AI talk to my patients on its own?

Not by default. The assistant can send messages, but it does so on a user's instruction. Fully automatic replies are possible; in healthcare our recommendation is to keep human approval in the loop — a wrong clinical implication costs more than the time saved.

Is my patient data used to train models?

No. The gateway returns data only for the question you ask, and MetoCRM does not use it for model training. Direct identifiers such as email and phone never enter the responses in the first place.

Which language model does it work with?

Because the protocol is standard, it is not tied to one model; it works with clients that support it. The choice of model stays yours.

What if it gets a report wrong?

The assistant reads numbers from the report endpoints rather than calculating them, so the figures match the CRM's own reports. The interpretation layer belongs to the AI and, like any interpretation, should be checked — for critical decisions, open the raw report too.

Does it cost extra?

The LLM gateway is enabled on request. Talk to sales for scope and terms.

Related pages

Conversion rates: which number tells you whatReports moduleSecurity

See it with your own data

In the demo we try the assistant against your own CRM questions — not a scripted scenario.

Request a demo Our security approach