Frequently asked questions
Answers on cost, privacy, self-hosting and governance.
Is ThinkShip ready to use today?
ThinkShip is under active development. The frontend is largely verified, the Rust backend is currently being hardened. A design-partner program supports early adoption — reach out via the contact page.
Why doesn't monitoring cost ongoing AI money?
Because waiting itself doesn't need language understanding. A deadline watch, a stock check, or a delay alert runs as a plain cron job or trigger — pure logic, 0 tokens. AI only switches in briefly once the workflow actually reports a hit (a deadline is coming up, stock has dropped below the safety threshold) to interpret the result or prepare an approval. So you pay AI cost for decisions, not for idle loops.
What does the AI itself cost?
Nothing that's included in the system price. Each tenant stores its own encrypted provider key or runs local models — there is no shared cloud account. See the pricing page for details.
Do I have to use a cloud provider?
No. ThinkShip treats cloud and locally run models as equal citizens. Data-residency routing can strictly limit sensitive tasks to local models only.
Can this be operated in a GDPR-compliant way?
ThinkShip runs self-hosted on your own infrastructure, with no shared cloud account — giving you the control that GDPR-compliant processing requires. Actual compliance depends on your own setup (hosting location, data-processing agreement with your chosen AI provider, etc.); we do not claim a blanket certification.
How quickly can the system be up and running?
An installer sets up the entire stack (database, identity management, messaging, model broker, backend, frontend) automatically and repeatably.
Do I need programming skills?
Not for day-to-day use — requests are written in plain language. Running your own infrastructure benefits from basic server/container operations knowledge.
How does approval for sensitive actions work?
Through a three-tier model: every action can be approved once, always, or never — configurable per topic, role, or individual action.
What happens when an agent hits its limits?
A repeated competence gap triggers a hiring event: a matching role is pulled from the talent registry or newly synthesized, and runs on probation with close compliance review at first.
How many industries does ThinkShip cover?
The core platform is built industry-agnostic. The industries page shows 20+ examples with concrete sample prompts — your own processes can be modeled via the role and talent registry.