All cases
FamelabSaaS / sales automation

An AI setter that books meetings and honours Article 50

Famelab's AI agent finds leads, scores them, sends personal LinkedIn invitations and books meetings straight into the calendar. With an AI disclosure nobody can switch off.

Art. 50
AI disclosure enforced by the system itself
24/7
in production, monitored on model behaviour and cost
0
settings that let you turn the disclosure off

The challenge

Famelab automates the entire lead generation process on LinkedIn. That is an application where two things meet that rarely go together: it has to run fully autonomously, and it has to be demonstrably well-behaved.

Autonomous, because the value lies in nobody having to sit with it. Well-behaved, because at the other end of every message is a real person who has the right to know they are dealing with AI. Since August 2026 that is no longer etiquette but an enforceable obligation under Article 50 of the EU AI Act.

The solution

The agent continuously finds and selects new leads based on preset criteria, assesses their relevance and scores them on likelihood. Personal invitations go out via LinkedIn, the most promising contacts automatically receive a meeting proposal, and the agent books those meetings straight into the calendar.

Details matter in a market where everyone knows everyone. Users work with exclusion lists to keep existing accounts and competitors out of scope, down to the guarantee that the AI never likes a competitor's post.

Compliance as a property of the software

The distinctive part of this case is not the agent. It is how the transparency obligation was solved.

We implemented a mandatory, non-bypassable AI disclosure field under Article 50. No setting you can switch off, no line in a user manual, no checkbox that happens to be ticked by default: the system does not let an outgoing message pass without the disclosure. Compliance is therefore not a promise about behaviour but a property of the product.

That is exactly what we mean by compliance as code. A lawyer can tell you that a disclosure is required. Whether it is actually there every single time is decided in the code.

Keeping it in production

An AI system doing outreach 24/7 fails differently from classic software. It does not fall over; it keeps going while it is already wrong. So alongside uptime we monitor model behaviour, cost per request, drift, and attempts to talk the system past its instructions through its own input.

That is why this system did not remain a pilot.

Results

  • A streamlined sales funnel with no manual work: from finding and scoring to a meeting in the calendar
  • Demonstrably compliant with the EU AI Act transparency obligation, enforced in code
  • AI-specific monitoring on model behaviour, cost and drift, alongside classic uptime monitoring

Related service

EU AI Act & compliance

Lawyers tell you what has to happen. We make sure it actually does: in the code, in the pipeline, in production. Transparency obligations, human oversight, logging and documentation: we build it in from day one, so compliance is not a brake but a given.

Request an AI Act quick scan

Next case

An enrichment agent that hands its own doubts to a human

Scobe · Sales & lead generation

Read the case