Privacy Policy

Crush Job is an AI product, and AI is not something your PC does by itself. To tailor a resume or transcribe an interview, the app has to send your text or your audio to an AI provider. That is the plain fact this policy starts with, because it is the one that matters most and the one a privacy policy is most tempted to bury.

What is true is narrower, and we think still worth choosing us for: with Cloud Sync off — the default — we store none of your content on our servers, your files and history live on your own machine, and the AI work we route on your behalf is forwarded to the provider without being kept.

What this policy covers

This policy covers both the crushjob.ai website and account (sign-up, billing, and support) and what the desktop app sends when you use it. crushjob.ai and the Crush Job desktop suite are operated by Crush Job LLC, a Washington limited liability company.

What the desktop app sends, and to whom

The app runs on your machine and keeps your profiles, resumes, and usage history there. Doing the AI work, however, means sending the relevant part of your data to the provider that performs it:

• Anthropic — Resume text, the job posting, and the answers you type into gap and strengthen questions — sent to be tailored, scored, and turned into battle cards and interview questions.
• AssemblyAI — Live audio, when you use Interview Copilot or Mock It!, for real-time transcription. Copilot captures two streams: your microphone and your PC's audio output, which contains the other participant's voice.
• Deepgram — Live audio for transcription, as above; and in Mock It!, the interviewer's question text for speech synthesis — unless you pick a Windows (SAPI) voice, which is local.

Those requests go through crushjob.ai. The app sends them to us, we forward them to the provider on our own provider accounts, and we bill the usage against your plan's credits. That is how the suite works for every linked account — there is no provider key for you to obtain, and the app has nowhere to enter one. Routed, not kept: we pass the request to the provider and record only token counts, model names, and operation labels — never your prompts, resumes, or transcripts.

Interview Copilot records the other side of a conversation. It captures your PC's audio output, which contains the other participant's voice, and sends it for transcription. Recording laws vary by jurisdiction — in some places every participant must consent before a conversation is recorded or transcribed. Obtaining that consent is your responsibility, not ours.

Cloud Sync

Cloud Sync is off by default, and it is a per-profile setting — you turn it on (or off) for each profile from the Hub's Account tab. Nothing in this section applies to you unless you turn it on.

When it is on, the app pushes structured copies of your own work to our servers so it can show up on your crushjob.ai dashboard: the profiles, applications, battle cards, mock-interview sessions and Copilot debriefs you create in the app.

It is a copy, not a move. The files on your own machine remain the original record — Cloud Sync never deletes or edits them, and turning it off simply stops the pushing.

Your interview transcripts and audio recordings are not part of it. A mock interview and a Copilot session each leave a full turn-by-turn transcript on your machine, and Mock It! can keep an audio recording as well. What Cloud Sync uploads is the structured summary beside them — the score, the debrief, the company and role. The transcript and the recording stay where they are.

You can delete the cloud copy at any time from the Hub's Account tab, and deleting your account removes it as well. We do not share synced data with third parties, and we do not use it to train AI models.

Crash and error diagnostics

If your device is linked to an account, the desktop app sends us a small report when it hits a crash or an unhandled error: the exception type, roughly where in our own code it happened, the app version, and your Windows build. Any text in that report is scrubbed of home-folder paths, email addresses, and long number sequences on your device before it is sent.

These reports are about the crash, not about you: they never include your resume text, interview transcripts, prompts, or any file contents. That makes them metadata, not content, so this does not change what we say above — unless you turn on Cloud Sync, your content is never written to our servers. The desktop app's crash reports go to our own servers, the same ones the rest of crushjob.ai runs on, and not to any third-party service. The website is a separate case, and the paragraphs after this one say how.

This is not a setting you can turn off on a linked device — we rely on crash reports from every linked install to find and fix issues across the app. If your install is not linked to an account, no crash reports are sent at all.

The website reports its errors to Sentry. When a page on crushjob.ai hits an error in your browser, or one of our own endpoints fails while serving you, we send a report to Sentry, an error-monitoring service acting as our processor. That report carries the error and where in our own code it happened, the page you were on, your browser and operating system, which build of the site you were running, and a short trail of the actions and requests that led up to it.

If you are signed in, the report also carries your account's id — the same identifier your account already has in our database — so we can tell one person hitting a problem ten times from ten people hitting it once. It does not carry your name or your email address. We do not record or replay your screen: Sentry offers that and we have it switched off. For errors raised on our servers rather than in your browser, we tell it to leave out request bodies and headers.

Information we collect

• Account data — your email address, and (if you sign in with a third-party account — Google, Microsoft, LinkedIn, GitHub, or Apple) the name and email address that provider shares with us, to create and secure your account. If you sign in with Apple and choose Hide My Email, we only ever see the private relay address Apple generates, never your real one.
• Contact data — the email address and message you submit through our contact form.
• Billing data — handled by our payment processor (Stripe). We store a customer reference and your subscription status, never your full card number.
• Product analytics — usage of the website (for example, which pages convert, and account-linked events like sign-up or a completed download) collected via PostHog, to improve the site. This only runs if you accept the cookie banner; declining blocks it entirely.
• Email engagement — what happened to a message we sent you: whether it was delivered, bounced, opened, had a link followed, or led you to unsubscribe. See the email section below for what is and is not recorded.

Cookies

We use a small number of cookies: session cookies required to keep you signed in, and — only if you accept the cookie banner shown on your first visit — analytics cookies from PostHog. You can change your mind at any time via the "Cookie preferences" link in the footer. We also use Cloudflare Turnstile, a bot-detection check on our forms, which does not track you across sites and runs regardless of your analytics choice since it's required to keep the forms working. One more cookie remembers, on this browser only, which sign-in method you used last — so the login page can point you back to it. That cookie holds the method's name only, never your email address.

What we do not collect

• AI provider API keys — there is no key of yours for us to collect. The suite reaches the providers listed above on our own provider accounts, and the app has no field for entering a key.
• Your resumes, prompts, or interview transcripts — with Cloud Sync off, we do not store them. They do pass through our servers on the way to the provider that performs the work, and are not retained there. If you turn Cloud Sync on, structured copies of your applications and interview sessions are stored so they can appear on your dashboard — see Cloud Sync above. Either way, the provider that performs the work does receive them — see what the desktop app sends above, and that provider's own policy for what they do with it.

How we use your information

• To provide the website, your account, and billing.
• To send you email about your account — the acknowledgement our contact form sends back, purchase receipts, credit-balance notices, and changes to our terms.
• To send you email you can switch off — recaps of your own activity, product updates, and job-hunt tips or offers. See the email section below.
• To understand and improve how the site performs.

Email we send, and how to stop it

We send four kinds of email. They differ in what they are for, and in whether you can turn them off.

• Account — purchase receipts, credit-balance notices, changes to our terms, a one-time sign-in link we email you, and the acknowledgement our contact form sends back. These come with having an account, so they carry no unsubscribe link.
• Activity — recaps of your own use of the suite, such as a monthly summary or a milestone you passed.
• Product updates — what we have shipped, and what is coming next.
• Tips and offers — job-hunt advice, help getting started, and occasional promotions.

The last three are on by default, and you can turn any of them off whenever you like. Every one of those messages carries an unsubscribe link that takes effect in a single click, and you can set all three at once on your email preferences page. Choosing to stop everything is recorded against your email address rather than your account, so it still holds if you sign up again later — and unlike the individual switches, it stops account email too.

Two providers deliver this mail. Zoho ZeptoMail carries account and activity email; Resend carries product updates and marketing, sent from a separate subdomain so that trouble with one kind of mail cannot affect delivery of the other. Each of them receives your email address, the message we wrote, and what became of it — delivered, bounced, or reported as spam. Neither receives your resume text, your prompts, or your interview audio.

We measure whether marketing email gets read. Product update and marketing messages contain a small invisible image, and their links point at crushjob.ai before forwarding you where they say they go. Between them, those record that a message was opened and that a link was followed. Both run on our own servers — no advertising network or analytics vendor is involved, and the provider that delivered the message does not see either one.

We keep those records deliberately thin. An open or a click is stored without your IP address and without your browser user-agent, and our database rejects a record that carries either. We record that a message was opened, not where you were when you opened it.

Account email carries neither of those. A receipt, a low-balance notice, or a change to our terms reaches you with no tracking image and no rewritten links, because that mail is not something we measure.

Service providers

To run the website and your account we use Supabase (database and auth), Stripe (payments), Zoho ZeptoMail (account and activity email), Resend (product-update and marketing email), PostHog (product analytics, opt-in), and Cloudflare Turnstile (bot protection on our forms). Each processes data only to provide their part of the service.

These are separate from the AI providers named above, which are the ones that receive your actual content. Both lists matter; only one of them is about your resume. The two email providers sit on this list rather than that one for exactly that reason: they are handed an address and a message we composed, never anything you wrote into the app.

Signing in with a third-party account — Google, Microsoft, LinkedIn, GitHub, or Apple — is also governed by that provider's own privacy policy. We receive only what the sign-in itself shares with us: your name and email address, nothing more.

Your rights

You can request access to, correction of, or deletion of your account data at any time by emailing support@crushjob.ai.

Contact

Questions about this policy? Email support@crushjob.ai. Crush Job LLC is based in Renton, Washington.