Short answers.

    Ownership, licence, models, data, and what your clients see.

    01Who owns the software?
    You do: the code generated for you, your data and your instance. The Product Printer software itself is licensed separately. Export everything whenever you like.
    02Is it open source?
    Not yet. At public launch it becomes source-available under FSL-1.1-ALv2, and every release becomes Apache 2.0 two years after publication. Until then, self-hosting is early access on request.
    03Which models does it use?
    Whichever you route to: local models, or frontier models with your subscription or keys, all on your orchestrator. The instance holds no model key.
    04Where is my data?
    Self-host: wherever you run it. Cloud: one private instance in Frankfurt, EU. AI calls go only to the models you route to; with local models, nothing leaves your premises.
    05What does "evidence" mean?
    A real test run, recorded and bound to the acceptance criterion it proves. Not a checkbox, not a chat transcript.
    06Who is behind it?
    LECAP UG (haftungsbeschränkt), Berlin. AvoDev, its agency, runs its client projects on the factory.
    07What do my clients see?
    Their own space, under your brand: requirements, open questions waiting for them, designs, worksteps in progress, the proof per criterion, the running preview and their change requests. Access is role-based: client, agency team, admin.
    08Can my clients change the software after launch?
    Yes. They describe the change in their space with AI help; it becomes a change request with acceptance criteria and an estimate. You approve it, agents build it, the proof runs before release.