AI disclaimer

Foundiry draws a hard line between the two, and it matters for how much you should rely on what you are looking at.

• Calculated. Runway, burn, milestone dates, dependency chains, what a hire costs you in months, when you have to start raising. These are pure functions of the figures you entered. They are deterministic, repeatable, free, and involve no model whatsoever. If the inputs are right, the output is right.
• Generated. Narrative prose, drafted sections, rewritten copy, suggested plans. These come from a language model. They are drafts.

Anything that costs credits is generated. Anything free is calculated. That is not a pricing coincidence — it is the same distinction seen from the billing side.

It is not advice, not a verified fact, and not a substitute for professional judgement. Language models produce fluent text that can be confidently wrong. Output may contain errors, invented specifics, or reasoning that does not survive contact with your actual situation.

Read it before you use it. Especially before you send it to an investor, a regulator, a lender or an employee.

• Generated output is structurally validated before it is stored, so a malformed response is rejected rather than saved as if it were fine.
• Nothing is generated without you approving it first, at a price shown up front.
• Missing information is shown as missing. The product does not fill a gap with a plausible-looking number — a blank you can see beats a guess you cannot.
• Generated content is versioned. You can see what changed and go back.

None of this makes model output correct. It makes it reviewable, which is a different and more achievable thing.

Only the context a specific approved action needs is sent to the provider that performs it, under terms prohibiting training on it. Files you upload are read once to extract facts you then confirm; they are not scored and your idea is not evaluated. See the privacy policy.

You are responsible for reviewing generated output before relying on it or passing it on. We are not liable for decisions taken on the basis of unreviewed output, and the liability limits in the terms of service apply.