What about my privacy?

Every accepted email goes through full redaction of personal information by our processing team before it enters any dataset — that's a core deliverable of the project, not an afterthought. You'll see the data-handling terms in full before submitting anything.

What counts as an acceptable email?

Original (not forwarded or replied), genuinely from your inbox, in the target language with a native-language subject line, and matching one of the open categories. The quality check screens for language, subject and category accuracy.

How many files can I submit?

As many as genuinely qualify. Categories have open volumes per language, and steady submitters of clean, valid files are the backbone of the project.

CB2026_003Active

The easiest gig on the platform: your old emails

Export genuine everyday emails in Hindi, Thai, Indonesian, Hebrew or Malay — receipts, OTPs, bills, promos — and get paid per accepted file.

Fully remote Work when you want Paid per accepted task

About the project

Somewhere in your inbox are hundreds of ordinary emails — order receipts, OTP messages, utility bills, calendar invites, promotions. To multilingual AI research, those routine messages are gold: real language, real formats, real variety.

This project collects genuine .eml files from native speakers of Hindi, Thai, Indonesian, Hebrew and Malay. You export messages from your own inbox that fit the listed categories, upload them as-is through our platform, and our team handles all redaction and privacy processing. Your only job is sourcing good files: original, correctly categorized, in the target language.

What you'll do

Find emails in your inbox matching the open categories (receipts, OTPs, bills, invites, promotions and more)
Export them in .eml format and upload through the platform
Pick the correct category and subcategory per file
Submit originals only — no forwards, replies, screenshots or fabrications

Who we're looking for

Native speakers of Hindi, Thai, Indonesian, Hebrew or Malay
People with an active personal inbox history in that language
Careful categorizers — accuracy is what gets files accepted

Requirements

A laptop and access to your own email account
Subject lines must be in the target language (3+ native-language words)
Every file must be genuine and your own
Paid per accepted file — volume adds up fast

How it works

1.Create your account
Sign up free with your email or Google. Takes about a minute.
2.Complete your profile
Tell us your languages, location and setup so we can match you.
3.You're automatically applied
Sign up from this page and your application to this project is submitted for you.
4.Our team reviews and onboards you
Applications are reviewed on a rolling basis and we follow up by email.

Frequently asked questions

What about my privacy?: Every accepted email goes through full redaction of personal information by our processing team before it enters any dataset — that's a core deliverable of the project, not an afterthought. You'll see the data-handling terms in full before submitting anything.
What counts as an acceptable email?: Original (not forwarded or replied), genuinely from your inbox, in the target language with a native-language subject line, and matching one of the open categories. The quality check screens for language, subject and category accuracy.
How many files can I submit?: As many as genuinely qualify. Categories have open volumes per language, and steady submitters of clean, valid files are the backbone of the project.

The work is already sitting in your inbox.

The work is already sitting in your inbox. Sign up to be automatically applied — the category list and export guide arrive with approval.

Already have an account? Sign in and apply

Browse all open projects