The rules we hold to.
Seven rules, each enforced by the product rather than promised in marketing, and how AI and security fit around them.
- T-01
Every claim cites your Record
Your Record is the experience you confirm: roles, wins, projects, stories, skills and credentials. Each item has a code such as R-03. Every résumé line, fit-brief finding and answer suggestion the studio drafts must cite one.
Before anything is shown, the studio checks it. A requirement must quote the posting word for word. A quoted excerpt must appear in the Record item it cites. A résumé line may not introduce a number, tool or credential that isn't in its sources. Anything that fails is removed, and you see how many lines were removed and why.
- T-02
Coverage, never a hiring probability
We show which requirements your Record supports, partly supports or doesn't show. We never estimate your chance of getting hired. Nobody can know that, and pretending to would push people toward the wrong roles.
- T-03
No mass auto-apply
The studio never submits applications for you. You decide where your name goes, and you mark a pursuit submitted yourself. Email you send from the studio goes one recipient at a time, up to 25 a day.
- T-04
Practice only
Rehearsals are for preparation. The studio won't feed you answers during a live interview, and the voice interviewer will decline if asked.
- T-05
You own your Record
Export everything as a file at any time from Settings. Delete your account and your studio is removed right away, along with any program invitations sent to your email; only security logs are kept, for up to two years. We don't sell data, and employers can't pay to rank you. Retention periods are listed in the privacy policy.
- T-06
Programs see only what you share
If you join a program, you choose what it can see: progress, documents, rehearsals and outcomes, each switched on separately and all off until you turn them on. Every time a coach or administrator opens something you share, it's written to an access log you can read, and so is each day your shared progress or work appears in their lists.
- T-07
General mode is labeled honestly
Without a field pack, fit briefs use general hiring judgment and say so. Field packs add expert depth for specific fields. We never present general advice as expert advice.
How AI is used
Fit briefs, résumé tailoring, letters and practice feedback are drafted by AI models (OpenAI or Anthropic, directly or through Vercel’s AI Gateway), then checked by the rules above. Neither we nor our AI providers use your Record to train models.
Voice calls are set up by our server, so no API key ever reaches your browser. The server writes the interviewer's instructions and reserves your minutes before the call starts. A supervisor process on our side joins every call, keeps those instructions fixed, ends the call when the reserved time runs out, and records the transcript your feedback is based on. If supervision can't be kept, the call ends. Audio isn't stored.
The Quick Scan on the home page uses no AI at all. It matches exact words in your browser, and nothing you paste is sent.
Security
Accounts support two-step verification. Every request is scoped to the signed-in account, actions are rate limited, uploads are size- and content-checked, and significant account events are written to a security log. Tokens for connected apps are encrypted, and disconnecting an app revokes its access. Payments are handled by Stripe; we never see your card number.
Reporting a vulnerability
Email security@careerstudioiq.com with steps to reproduce. Please don't access other people's data, degrade the service or disclose the issue publicly until we've fixed it. We'll acknowledge reports within three business days and won't pursue legal action for good-faith research that follows these rules. Our security.txt lists the same contact.