Privacy

A published developer profile is public, because being found is the point of having one. Everything else — drafts, saved lists, applications and messages — is visible only to the people directly involved, and that is enforced by the database rather than by the application remembering to check.

Once you publish a developer profile, anyone can see:

• Your headline, bio, skills, rates, availability, timezone and years of experience
• Your portfolio items, including any links and images you add to them
• The work experience and education you add
• Any website, GitHub or LinkedIn links you provide

The same applies to a job post once it is open: its title, description, budget range, required skills and the company name behind it.

Search engines index these pages. That is deliberate — a profile nobody can find does not do its job — but it means you should treat anything you put on a published profile as public, permanently, including after you unpublish it.

None of the following is visible to anyone but you and, where relevant, the other party:

• An unpublished profile or a draft job post — only you can load it
• Your email address. It is never shown on a profile or a job post
• Your saved jobs and saved developers. These are one-directional: a developer cannot see which clients saved them, and a client cannot see who saved their job
• Applications, cover notes and their status history — the applicant and the job owner only
• Messages — the two people in the conversation only

Messages are plain text and are stored in full. Both participants can read the whole thread; nobody else can, including us in normal operation, though as the people who run the database we can technically read it and would if legally compelled or when investigating a report.

Blocking is mutual and silent. If you block someone, neither of you can write to the other, and they are not told that it happened. We consider this a privacy feature: someone who learns they have been blocked tends to make a second account.

Reports are one-way. The person you report cannot see that you reported them. A person reads reports; there is no automated action and no fixed turnaround, so blocking is the control that takes effect immediately. When a conversation or a message is reported, the person handling it reads the latest messages in that conversation, because a report cannot be judged without them.

This is a small site and it stands on other people’s infrastructure:

• Supabase — the database, authentication and file storage. Everything described on this page physically lives there
• Vercel — hosting. Receives the usual request metadata, including your IP address, in order to serve pages
• Google and GitHub — only if you choose to sign in with them. They tell us your email address, name and avatar; we do not receive your password and we do not ask either service for anything beyond your basic profile
• Resend — sends transactional email. Receives your email address and the contents of that email
• Google (Gemini API) — powers matching. See below, because this one carries a condition worth reading rather than a line in a list
• Sentry — error tracking, when it is switched on: if something breaks, it receives the technical error, the page path and the browser, so the problem can be fixed. Never your email address, messages, form contents, cookies or anything you typed

There is no analytics, no advertising and no tracking on this site. The only cookies set are the ones that keep you signed in. Error reports (Sentry, above) are about the software failing, not about you, and set no cookies.

Matching works by turning text into a numerical representation and comparing it, then asking a language model to judge the shortlist and write the one-line reason you see beside each match. Both steps run on Google’s Gemini API.

What is sent: job titles and descriptions, and developer headlines, bios, portfolio text and certification names. Only from published profiles and open jobs — the things already visible to anyone on the internet and indexed by search engines.

What is never sent: email addresses, messages, applications and cover notes, saved lists, drafts, or any unpublished profile. Matching reads what you chose to publish and nothing else.

One exception you start yourself: importing a resume. If you use “Import from resume”, we read the text of the PDF you upload, remove email addresses and phone numbers, and send the rest to Gemini to sort into profile sections. This happens only when you press the button, after you confirm it, and it can include text from a profile you have not published yet. The file itself is never stored, and nothing is added to your profile until you choose what to keep. The free-tier note below applies to this text too.

We use Google’s free tier, and Google states that content sent to it may be used to improve their products. In practice that means the public text described above, and the text of a resume you import, may be used in training. Everything else private to you — your messages, your applications, your email — is never part of it, because it is never sent.

If matching is unavailable, nothing else breaks: profiles, jobs, applications and messages all work exactly as they do now. You would simply see “no strong matches yet” instead of a ranked list.

Only email tied to something you did or that was sent to you: confirming your address, a welcome when you first sign in, resetting a password, a new application on your job, a change to an application you submitted, and a new message (at most one email per conversation until you have read it, without the message itself). There is no marketing email, so there is no marketing list to unsubscribe from.

You can delete your account, and it genuinely deletes. Your profile, portfolio, uploads, job posts, applications, saved lists and blocks are removed.

Conversations are the deliberate exception. A conversation has two owners. If it were destroyed when one person left, the other would lose their record of an exchange they may still need. So the thread survives with your side tombstoned: your name is detached from it, and the text of your messages is replaced with a note saying the account was deleted. The other participant keeps the shape of the conversation and their own words; they cannot recover yours.

Data stays until you delete it or delete your account. We do not currently expire old jobs, applications or messages automatically. A resume you import is not kept at all: it is read once and discarded.

You can see and edit everything on your profile from your dashboard, download a copy of your data from Settings → Account, and delete your account there at any time.

This is an independently built product, not a company with a compliance department. There is no round-the-clock moderation: reports go to a queue that one person works through, and blocking is what takes effect at once. No system is perfectly secure, and we are not going to pretend otherwise. If you find a problem with how data is handled here, tell us and it will get fixed.

If this policy changes in a way that affects what is visible to whom, the date at the top changes with it.