Affiliate.Careers

Why we exist

Affiliate marketing and traffic arbitrage is a closed market. By our estimate, around 80% of jobs in the industry are never posted on standard job boards or LinkedIn. Employers find people through Telegram channels, personal networks, referrals, and private chats. This model has one big problem: it works well for people who are already inside, and almost not at all for everyone else.

If you're a junior — you don't know where to write or what even exists on the market. If you're middle — you don't know what a fair salary is for your experience. If you're senior — you miss strong teams just because you follow the wrong channel. And if you're an employer — your job gets lost among dozens of similar ones posted the same day.

Affiliate.Careers tries to solve this mechanically: gather all the scattered jobs in one place, normalize them into one format, deduplicate, give proper filters, and leave the user alone. No registration, no lead-gen, no "leave us your contact — we'll call you back".

Team

Affiliate.Careers is run by an independent editorial team specializing in the CPA industry. We are not affiliated with any employer in the catalog and do not receive commissions from any side.

The core team is two developers and an editor-analyst with performance marketing experience since 2018. We deliberately do not list names: the project is meant as a service, not a personal brand. If you need to reach us, the editorial email is faster than searching for a specific person.

We are an independent editorial team and do not work with any employer from the catalog on a paid or barter basis. This is a deliberate stance: a catalog loses its value the moment it acquires a commercial interest. Contact for editorial, press and proposals — [email protected].

What we do technically

Building the catalog is a fully automated pipeline. We parse public Telegram channels 24/7 via the Telethon library. Each message goes through several filtering stages: first, obvious noise is filtered out (candidate resumes, news, ads, congratulations), then job signals are extracted from the text — role, company, work format, salary, contacts for applying.

Next is deduplication by content: we compute a canonical hash from key fields and collapse the same job posted in ten channels into one record with a note about all sources. Then a classifier assigns tags: vertical (gambling, betting, dating, nutra, crypto, finance, white), role (media buyer, affiliate manager, farmer, team lead, designer), seniority, and work format.

Borderline cases — about 5-8% of jobs where automation is not sure — go to AI moderation via Claude Sonnet. The model checks whether this is really a job, not a promo or an attempt to collect resumes under the guise of hiring, and extracts data from poorly structured text when needed. Each job lives 30 days: if it never reappears in a source during that time, it auto-archives. Methodology details are on /methodology.

What we don't do

• We don't take commissions — neither from candidates for applying, nor from employers for posting. The project has no tariff plan and no "premium placement".
• We don't post "bait jobs" — i.e. fake postings that many job boards put up only to collect resumes and resell them to recruiters.
• We don't sell your data — we don't collect it. There is no signup, no account, no resume upload on the site.
• We don't replace employer contacts with our own. All "apply" links go directly to where the original posting said: HR Telegram, company website form, email.
• We don't let obvious scam jobs through. Postings without descriptions, fantasy salaries for junior roles without requirements, schemes with "entry fees" — Claude Sonnet cuts those at moderation, and borderline cases are reviewed manually by an editor.

How to use the catalog

• Home — top fresh jobs and FAQ for first-time visitors.
• /jobs — full catalog with filters by role, vertical, work format, salary.
• /companies — employer cards with each employer's job history.
• Topical hubs: /jobs/role/{X}, /jobs/vertical/{X}, /jobs/format/remote — collect jobs by a specific slice.
• Markdown versions: /job/{slug}.md — clean markdown without HTML for AI assistants.
• Machine feeds: /feeds/jobs.rss, /feeds/jobs.json, /vacancy-feed.xml — RSS, Yandex Vacancy Feed, JSON Feed for integrations.

Contact

If an employer wants their job removed — write to [email protected], we'll remove it within 24 hours after verifying the request.

If you spot a scam posting, a duplicate, or a misclassified job — write to the same address. Every message is read by a real editor.

For editorial, media and partner questions — also [email protected].