Privacy

The short version

Good Roles Only watches job postings for you and emails the ones worth your evening. To do that we need your email address and a description of the role you want. That is nearly all of it.

We do not sell anything to anyone, we run no advertising, and there is no analytics on this site — no page tracking, no third-party scripts, no profile of you anywhere.

What we hold

Everything in this list is something you typed, uploaded or asked for.

• Your email address, so we can sign you in and send you the digest.
• The description of the role you want. Every edit is kept as a new version, because a score has to record which description it was scored against.
• Your searches — up to 3 of them — and when you want to hear from us.
• Every posting we have scored for you, its score, and the reasoning behind it.
• Whether you gave a match a thumbs up or down, and whether you marked it applied.
• A time zone, so a digest at 11:00 arrives at 11:00 where you are.

Your resume, if you upload one

The file is never stored. It is read once, on its way through, to write a first draft of your description — and then it is gone. There is no bucket, no attachment, and nothing to leak later.

The draft appears in the box for you to read and edit, and nothing is saved until you press the button. The model is instructed to leave out your name, address, phone number, photograph, date of birth, nationality, and anything about your health or family — what it keeps is employers, schools, job titles and tools.

We do not claim to strip anything out of the file itself. Nothing filters it: it goes to the AI service as you uploaded it, which is why we ask you to tick a box first, and why the draft is yours to read before it becomes anything.

What the AI does and does not do

Two things. It reads each job posting against your description and scores it, and — only if you upload a resume — it writes that first draft.

It is not trained on anything of yours. Anthropic's commercial terms say they may not train models on customer content, and this runs under those terms.

The thumbs are recorded and nothing learns from them automatically. They are there so a person can look at whether the matching is any good; no part of the system reads them back.

How long any of it lasts

Your account's rows last until you ask us to delete them. Scores accumulate — the history page is the point of keeping them.

At the AI provider, the scoring calls are held for up to 29 days and then expire; the drafting call is not stored at all. When you ask us to delete your account we delete those held calls too.

Emails we have already sent are in your inbox and are not ours to take back. The companies above keep their own operational logs on their own schedules, which we do not control.

Deleting everything

Email codewcj@gmail.com and say so. We delete your account, your searches, your descriptions, everything scored for you, your ratings, and the sign-in record itself. Within 30 days, and usually the same day.

There is no button for this yet, which is why the address is here rather than a link. The same address is where to write to ask what we hold about you, or to have something corrected.

Two things survive on purpose, and neither names you: the job postings themselves, which are public and shared by everybody, and the search terms, which are shared between people looking for the same thing.

Cookies

Two, and both are the site working. One keeps you signed in. One remembers whether you collapsed the sidebar.

There is no analytics cookie, no advertising cookie, and nothing from a third party.

Who else is involved

Running this needs six other companies. Each one is listed with what it does and what of yours reaches it.

Supabase

The database, and the sign-in link.

Your email address, the descriptions you write, your searches, and every posting we have scored for you. Hosted in the United States.

Anthropic

The AI that reads postings, and drafts a description from a resume.

Your description and the text of the postings being scored against it. If you upload a resume, the resume itself, once.

Vercel

Runs this website.

Whatever your browser sends, including a resume on its way to be read. Nothing of yours is stored there.

Resend

Sends the email.

Your email address, and the matches in each digest.

Railway

Runs the scheduled job that does the searching and scoring.

The same database rows, from the other side. Nothing you send it directly.

Apify

Fetches the job postings from LinkedIn.

The search terms — job titles, locations, filters — with nothing identifying you attached. Searches that mean the same thing are shared, so this is not per-person.

The legal part

We hold and use this to provide the service you asked for — under GDPR that is Article 6(1)(b), performance of a contract. A resume can carry special category data, which needs explicit consent on top of that: the box you tick before uploading is that consent, and it names what it covers.

You can ask for a copy of what we hold, ask for it to be corrected, ask for it to be deleted, withdraw consent, or complain to your data protection authority. Everything but the last one starts at codewcj@gmail.com.

Our servers and database are in the United States, so using this means your data is handled there.

Good Roles Only is an independent project rather than a company, and codewcj@gmail.com reaches the person who runs it.