Privacy policy
This page says what trakappli keeps, why, who else may see part of it, and when it goes away. Every section has a fixed anchor, so it can be linked directly.
What we collect
- Applications — the jobs you track, their status history, notes, reminders, events and tags.
- Profiles — the facts you give about yourself: skills, experience, education, languages.
- Documents — the CVs and cover letters you upload, and the text extracted from them.
- Captures — the job postings you save from a job board with the browser extension or the mobile share sheet.
- Your account — your email address, your sign-in method, your language, your theme and your credit balance.
AI providers
Some features (extracting a CV, scoring a posting against a profile, drafting text) call an AI model hosted by a third party. The models used are free models reached through OpenRouter, the gateway, with one direct free provider as the fallback when OpenRouter's quota runs out.
What such a provider may receive:
- the redacted text of your CV (see redaction);
- the text of the job posting;
- the profile facts the feature needs.
What it never receives: the original file you uploaded. Files stay in our private storage.
Redaction
Before any text is sent to a provider, your phone number, postal address, exact date of birth and email address are removed, and your name becomes initials: two letters per part, joined by _. Mohamed Maatar becomes Mo_Ma. What was redacted is recorded with the call.
Benchmarks
Aggregate comparisons — response rates by board, by role family and by job title — are one of the purposes of processing. They are a calculated view of the data you already entrust to the service, so there is no separate consent switch. A figure is never shown below 20 responses; below that floor the page says the statistics will be shown later. A benchmark never names or identifies anyone.
Shared postings
A job is stored once, as one posting, however many people save it. Your corrections to a posting are yours and are kept per user. Nothing tells you, or anyone else, who else applied to the same job.
Retention
Deleting something marks it deleted at once; it is erased for good after 90 days. Deleting your account deletes everything that belongs to it on the same terms.
Financial records
Records of purchases and refunds are kept after an account is deleted, because accounting law requires them, but they are severed from any identifier that could lead back to you.
Training material
Structural material — the shape of documents and postings, the corrections made to extractions — may be kept on our own servers to improve extraction, with the link to your account severed. CV prose and original files are never kept for this.
Captures
The raw HTML of a captured page is kept, compressed, so a posting can be re-read later. Screenshots from the mobile capture are kept for 90 days, then erased.
The browser extension
The extension reads only the page you click it on: one of the five supported job boards (LinkedIn, Indeed, Glassdoor, Welcome to the Jungle, HelloWork), or any other page when you ask it to. Our server never fetches a job board itself.
Hosting and processors
Text extraction (reading a CV or a posting into fields) runs on a server we own, hosted at home (a homelab). It is reached from the internet through a Cloudflare Tunnel, so Cloudflare carries that traffic, encrypted, between the application and the homelab. The homelab keeps no copy of what it reads. When the homelab is down, you can still save without AI.
Storage
Your files sit in private storage. They are reached only through short-lived signed links, made for you when you open a file.
Errors and analytics
Application errors are recorded by an error tracker we host ourselves. There is no third-party analytics and no advertising tracker.
Your rights
You can export your data and delete your account from the settings at any time. See the terms for the rest.