Summary
What most people want to know, in plain language:
- You have an account. Sign in with an email and password, or with Google. That account owns everything you put in.
- Your resume text and your interview recordings are sent to Google's Gemini models. That is not a side effect: it is how the reviews, tailored resumes and debriefs get written. They are sent without your name attached to the request, but the resume text itself contains whatever your resume says.
- Interview audio is deleted once your transcript is written. We keep the transcript and what is built from it, not the recording. Deleting an interview round deletes those too.
- An interview has another person in it. Their voice, their words and often their name end up stored with yours. Asking them first is your responsibility, and Prept reminds you every time.
- No advertising, no data brokers, no selling. The app carries no analytics or tracking of any kind. This marketing website does use Google Analytics; see Cookies.
- You can delete things yourself (any resume version, application, interview round or transcript), and you can ask us to delete the whole account.
If anything here is unclear, or you want to know exactly what is stored for you, email hello@prept.fyi. A real human will reply.
Who we are
Prept is a product of Sarosh Mirza Consulting Ltd., a company registered in British Columbia, Canada (registration number BC1486183). That company is the data controller for everything described on this page.
The app runs at app.prept.fyi; this site, prept.fyi, is where the guides and the rest of the writing live. Contact for anything on this page: hello@prept.fyi.
What we collect
Your account
Your email address and your name. If you sign up with a password, we store it hashed with scrypt and never in a form we can read. If you sign in with Google, Google tells us your email address, your name, an identifier for your Google account and the web address of your Google profile picture, and we store those four things, and none of Google's access tokens, which are discarded as soon as the sign-in completes. We keep the picture's address only to show your picture in the app's header: your browser loads it from Google, it is updated or removed each time you sign in with Google, and it is deleted with your account.
We also keep the housekeeping that comes with an account: when you signed up, when you last signed in, how many times you have, and whether the account is active.
When you create an account we also note how you first found us: the campaign tag on the link you followed, the site the click came from (its name only, such as linkedin.com, not the page), the app whose browser opened the page if there was one, and the first of our pages you opened. Your browser holds this for up to 30 days from that first visit and hands it over once, when the account is created. It is kept with your account, so we know which channels bring people to Prept.
Your resume
The file you upload and the plain text taken out of it. The text extraction happens on our own server. The document file itself is never sent to an AI model. Every version you upload is kept as its own version, so you can go back, and none are overwritten. Because a resume header usually carries them, this normally includes your phone number, location, and links to LinkedIn, GitHub or a personal site.
It also includes anything you type when Prept asks you to fill a gap: the figure behind a claim, the detail a line is missing.
Applications
The job postings you paste or link, the company and role, your own notes, and everything Prept generates from them: the fit check, the tailored resume, the prep sheet.
Interviews
For each recorded round: the transcript, the debrief, the scorecard, the quotes pulled from what was said, the names of interviewers if you enter them, and when it happened. See Recordings and other people.
Your career facts
The record Prept builds from the above (employers, job titles, dates, results), each stored with the sentence it came from, so every line can be traced back to something you confirmed.
Application details you choose to enter
Work authorisation, sponsorship needs, relocation and office preferences, your notice period and salary expectation if you give them, and optionally the diversity questions application forms ask. See Sensitive details.
The browser extension
If you install the Prept extension for Chrome, it reads only the web page in front of you, and only while its panel is open. If you let it read pages as you open them, it reads each page you open while the panel is open, to show you the job; it saves a job to your account only when you press Save or Apply. From a job posting it keeps the page's address and the posting's title, company, location, salary if stated, and text, saved to your account as a job. When you press Apply it fills the form in your browser with your name, contact details and links, your current job title and employer from your confirmed record, and your application details above, and attaches your resume. What it filled is not sent to us; we keep only counts (how many fields were filled, and how many you changed afterwards), which resume went with the application, and when you marked it applied. After it fills a form it notices when you press the employer's Submit button and looks for the employer's confirmation, so it can mark the job applied (you can turn this off in its settings, or undo it). It never presses Submit itself, and never reads other tabs, your browsing history or your LinkedIn messages. Of what you type into a form, it reads only your answers to the questions it left for you, once, when you press Submit, and shows them to you in its panel. It keeps an answer only if you tick it, as one of your Application answers in Settings, where you can change or remove it, and it fills the same question on the next form with it, outlined in blue for you to check. It never keeps a password, an identity, tax or bank number, a date of birth, or an answer to the voluntary self-identification questions, and these answers are never sent to an AI model. When it cannot read a page, find a form, or fill one well, it tells us so we can fix it: the page's address without anything after a "?", what it read from the posting (title, company, location, salary and the length of the text), and the names of the form's fields, which ones it filled and which it left for you, never what is in them. For a page it could not read at all it sends only the site's name. If you press "Report this form", it sends the same. It signs in with a token kept in the extension's own storage in your browser; you can disconnect any browser from Settings in the app, which deletes its token. The app asks the extension only whether it is installed in that browser, signed in and pinned to the toolbar, so it can show you the right next step. The first time the app sees you on a computer that could install the extension, we note the date, so that the one email we send about it, two days after you confirm your facts, goes only to people who can use it; if you ask the app to email you the link instead, that is the one email.
The free tools, without an account
If you use the free ATS checker, the free resume optimizer, free job match or the free cover letter generator on this site, we keep the resume you upload, the text taken out of it, any job description you paste, anything you write about why you want the job, and the report, result or letter, on our server. We keep them so you can pick up where you left off, so your resume comes with you if you create an account, and so we can see how the tools are used and improve them. They are kept until you ask us to delete them: every report has a "Delete my resume now" link that removes them at once, or you can email hello@prept.fyi. We also keep a random identifier stored in your browser, so uploads from the same browser can be counted together. We do not use the contact details in your resume to contact you or for marketing. The text of the resume, and the job description if you paste one, are sent to Google's Gemini model for the AI parts of the report; Google does not use them to improve its products. The score and the other checks come from our own rules.
The free interview question generator takes no resume. We keep the job description you paste and the questions we made from it, on the same terms: until you ask us to delete them, with a "Delete this now" link on every result, and carried into your account if you create one. The job description is sent to Google's Gemini model to write the guidance.
To stop the free tools being overused, our server counts uses by internet address, for an hour and for a day, in memory only, and keeps no record of the address. To stop scripts using them, the tools ask Cloudflare Turnstile to check that a person is using them. Turnstile looks at signals from your browser and your internet address to tell a person from a script, usually without asking you anything, and tells us only whether the check passed.
What our AI models were sent, and what they wrote
Every request we send to an AI model on your behalf, and its full answer as it came back, before our own checks change or drop any of it. And every result we showed you, as you saw it, including earlier versions of things you made again (a tailored resume, prep, your LinkedIn profile) and previews you never saved. Audio is not part of this: a request that carried audio is kept with the audio's size in its place. We keep these so we can check the results, find what went wrong when something does, and improve the product. They are kept with your account, or for the free tools with that upload.
Usage records
Which features you ran and when, how much model time each one used, and what it cost us to serve. These carry no content. They are counts, so we know what the service costs to run and can tell you what your plan has left.
Recordings and other people
This is the part that deserves more than a sentence, because an interview recording is not only about you.
What is stored, and for how long
When you record a round, the audio is written to our server as you go so it can be transcribed. Once the transcript is written, the audio is deleted. What stays is the transcript and what Prept builds from it, until you delete it.
The other person
An interview recording contains an interviewer's voice and words, and the transcript, debrief and quotes carry them too. That person agreed to talk to you; they did not sign up for Prept. So:
- Prept asks you to confirm, every single time, that everyone on the call knows they are being recorded. You are telling us you have their consent, and we take you at your word: we cannot verify it.
- Recording law differs by country and by state, and some places require every person on a call to agree. Following the law where you and they are is your responsibility.
- Their words are never used to build anything except your own debrief. They are not used for training, research, or anything shared with anyone else.
- If an interviewer contacts us about an interview they appear in, we will find the transcript and delete it. They can write to hello@prept.fyi.
Your voice
Your own recorded voice is personal data, and in some places it has specific protection of its own. We use it to produce your transcript and nothing else, and it is deleted as soon as the transcript is written.
Sensitive details
Application forms routinely ask about gender, ethnicity, veteran status and disability. Prept can hold your answers so you are not retyping them into every form, and they are optional. Leave them blank and nothing else changes.
Where they are concerned, the app is built to move them in one direction only: into a form you have reviewed. They are never included in anything sent to an AI model, never used to write or score anything, and never used to decide what Prept shows you. If you would rather not store them at all, clear the fields and they are gone.
Where your data goes
There are four places outside our own server that any of this reaches, and no others.
Google: Gemini models
This is the one that matters, because it is the product. Sent to Google's Gemini API:
- Interview audio, in slices, to produce the transcript, covering both your microphone and the other side of the call.
- Your resume text, to parse it, to write your review, and to write tailored resumes. The contact details in your resume header go with it, because they are part of the text.
- Interview transcripts in full, to write the debrief and pull out what you claimed.
- Job postings, and the confirmed facts about your career, to write fit checks, tailored resumes, prep sheets and your story.
These requests carry no account identifier, no name of ours for you, and no session: only the text or audio and the instruction. We use the paid Gemini API, whose terms state that content submitted through it is not used to train Google's models. A request that fails or produces something that does not pass our own checks is retried, which means the same content may be sent more than once for one action.
Google: sign-in
If you use "Continue with Google", Google handles the sign-in and tells us the three things listed above. Nothing about your use of Prept goes back to Google through this.
Job sites
When you paste a link to a posting, our server fetches that page, not your browser. The site sees a request from us, with no cookies and nothing about you attached.
Cloudflare: the free tools' human check
On the free tools on this site, Cloudflare Turnstile checks that a person, not a script, is using them. Cloudflare sees your browser's signals and internet address for that check, under its own privacy policy, and tells us only whether it passed. It is not used in the app.
Our host
The app and its data run on Railway, which provides the servers and storage. They hold the data on our behalf and do not use it for anything of their own.
That is the complete list. There is no analytics in the app, no error-reporting service, no email provider, no advertising network, no data broker, and today no payment processor.
How we use it
- To do the work you asked for: read the resume, check the fit, write the tailored version, transcribe the round, write the debrief.
- To keep your record straight, so that what you confirmed in one step is there in the next.
- To run the service: knowing what a feature costs to serve, what your plan has left, and whether something is broken.
- To reply to you when you get in touch.
We do not use your content to train models, ours or anyone else's. We do not use it to write marketing, articles or examples, even anonymised.
What we never do
- Sell your data, or share it with advertisers or data brokers. There is nothing to sell it through: the app has no advertising or tracking in it at all.
- Use your resume, transcripts or debriefs to train a model.
- Show your content to another user. Everything is scoped to the account that created it.
- Send your resume file to an AI model. Only the text read out of it is sent.
- Use the diversity answers you store for anything but the form you are filling.
We may disclose data if the law requires it, and then only the minimum required, and only when we have to.
Storage and security
Your data is separated by account, by construction rather than by convention: resumes and recordings sit under a path belonging to your account, and another account's path does not exist for you. Every database read and write runs inside your account's context, which is what keeps one person's record out of another's.
Passwords are hashed with scrypt and salted per account, so nobody here can read yours, and a reset means setting a new one. Your session cookie is signed, marked HttpOnly so page scripts cannot read it, and sent only over HTTPS in production. Signing out revokes the session on the server, not only in your browser.
Administrative access to the systems your data sits on is limited to the people who need it to run the service. The honest statement is that your content is technically reachable by us, not that it is sealed away. That is true of every service that can restore your account when something breaks. Our internal tools show accounts, usage and cost, not the contents of anyone's resumes or transcripts, and every act of granting access to an account is written to an audit log with who did it, to whom, and why.
No system is perfectly secure, and we would rather say that than imply otherwise.
How long we keep it
Your content is kept until you delete it or ask us to close the account. There is no automatic expiry on resumes, applications or transcripts, by design: this is a record you build over a job search, and a job search can run long.
Some things do expire on their own: interview audio and any working files made while transcribing it, which are deleted as soon as the transcript is written, and your sign-in session, after 30 days.
What our AI models were sent and wrote, and the results you were shown, are kept until the account is closed.
Usage and cost records (counts, not content) are kept as the accounting record of what the service did.
Deleting your data
Yourself, at any time
From inside the app you can delete any resume version, any application, any interview round, and any transcript. An application takes its rounds, transcripts, debriefs and claims with it.
The whole account
Email hello@prept.fyi from the address you signed up with, and we will delete the account and everything belonging to it (database records, resume files, transcripts, and what our AI models were sent and wrote) within 30 days, and confirm when it is done. We are building a way for you to do this yourself without asking; until it is there, one email is the way.
Your rights
Depending on where you live, you may have the right to see what we hold about you, correct it, delete it, get a copy in a portable form, object to a particular use, or withdraw consent. Regardless of where you live, we will honour all of those.
Email hello@prept.fyi and we will answer within 30 days. We will not charge you and we will not make the service worse for you for having asked.
We process your data to provide a service you asked for and, where a request depends on something optional like the diversity answers, because you chose to give it, which you can withdraw at any time by clearing the fields. If you are in the UK or EU and you think we have handled something badly, you can also complain to your national data-protection authority, though we would rather you told us first.
The servers and the AI processing run in the United States. If you are somewhere else, your data is transferred there to provide the service.
Children
Prept is for people looking for work and is not directed at children. We do not knowingly collect data from anyone under 16. If you believe a child has an account, email us and we will remove it.
Changes to this policy
We will update this page when what the service does changes. The date at the top is the honest one: it moves when the words do. For anything material, like a new company receiving your data or a change in what we keep, we will tell account holders directly rather than quietly editing this page.
Contact
Questions about your privacy, a request about your data, or a recording you appear in: hello@prept.fyi.
Terms of Use
By using Prept you agree to these terms. They are meant to be read, so they are short.
Who you are dealing with
Prept is a product of Sarosh Mirza Consulting Ltd., a company registered in British Columbia, Canada, registration number BC1486183. "We" and "us" on this page mean that company. It is the party you are contracting with, the party that takes your payment, and the party responsible for the service.
Your account
You need an account to use Prept, and you are responsible for what happens under it. Keep your password to yourself, and tell us if you think someone else is in there. One account is for one person. We can suspend an account that is being used to break these terms, and we will tell you why.
Your content is yours
Your resume, your recordings, your notes and everything built from them belong to you. You give us only the permission needed to run the service on them: to store them, to process them, and to send them to the AI models described above so the thing you asked for gets made. That permission ends when you delete the content or close the account.
Recording
You may only record a call you are lawfully entitled to record, and you confirm each time that everyone on it knows. Recording law is local and it is on you to follow it. We will remove any transcript of a recording made without consent, on request from anyone in it.
What Prept writes
Prept will not invent a number, a skill or a result, and nothing goes on your resume that you have not confirmed. It is still software writing text: read what it produces before you send it to an employer. The words go out under your name, and the responsibility for them is yours.
No guarantee of outcomes
Prept is a tool for doing a job search properly. It cannot promise you interviews, offers, or any particular result. Anyone promising that is selling something else.
Acceptable use
Do not try to break the service, scrape it, resell it, reverse-engineer it into a competing product, use it to impersonate someone, or use it for anything illegal. Do not upload someone else's resume as your own, and do not record people you have no right to record.
Our content
The software, the prompts behind it, the guides on this site and the way Prept structures its work are ours. Use them inside the product; do not redistribute them as your own.
Price and billing
Prept is free while it is in early access. If and when paid plans start, we will say so clearly before anything is charged, and no existing account will be billed without having agreed to it first.
Payments are processed by Stripe. Your card details go to Stripe and are never held by us. We are told that a payment succeeded, which plan it was for, and when the period ends, and nothing else. Prices are in US dollars, and any sales tax or GST that applies where you are is added at checkout.
A subscription renews on its own until you cancel it. You can cancel at any time from the billing page inside the app, and access continues to the end of the period you have already paid for. Cancelling does not cut you off mid-month.
Refunds
If you have not used the paid features, ask within 14 days and we will refund you in full. No argument and no form to fill in: email hello@prept.fyi from the address on the account.
Once you have used them, the purchase is not refundable. Using them means any of these:
- A resume tailored to a posting
- A job posting read and analysed, or a fit check run against it
- A prep sheet, a story, or a resume review generated
- An interview transcribed, or a debrief written from one
This is not a trick to keep your money. Each of those runs costs us real money the moment you ask for it. The work is done by AI models we pay for per use, and it cannot be handed back. That is also why the free tier exists: it is there so you can see exactly what the product does, on your own resume and your own interview, before paying anything.
Two things that are never charged, and so never need refunding: a run that fails, and reading something you already produced. Both are free on every plan, always.
If something went wrong on our side (you were charged twice, a plan did not activate, the product did something it should not have), that is not a refund question and the policy above does not apply. Tell us and we will put it right.
Availability
Prept is provided as is, without warranties. We work to keep it running and to not lose your data, but no service is immune to failure, so keep your own copy of anything you cannot afford to lose. To the maximum extent the law allows, we are not liable for indirect or consequential losses arising from your use of Prept, including a job you did not get.
Ending it
You can stop using Prept and ask us to delete your account at any time. If we ever shut the service down, we will give you notice and time to take your data out first.
Changes
If we materially change these terms, we will post the new version here and tell account holders. Continuing to use Prept after that means you accept them.
Contact
Questions about these terms: hello@prept.fyi