If you've done annotation or AI training work elsewhere, you've probably run into the industry's common frustrations: support tickets that go unanswered, accounts deactivated without explanation, unclear pay, and hours of unpaid assessments. We built HumanSignal Labs to work differently. Here's what you can expect from us.
You'll always know what a project pays — before you start
Pay rates are disclosed when you're accepted to a project, before you begin any paid task. Rates vary by project complexity and required expertise, but you'll never be asked to start production work without knowing what it pays. Your earnings accumulate in your wallet, where you can see your balance and request payouts at any time.
Real humans answer your questions
Every contributor has access to:
- A support team that reads and responds to every ticket — see Contact support
- A contributor community on Discord, where you can ask questions and connect with other contributors
- Project-specific channels for guideline questions on active projects
Managed projects, not a free-for-all
HumanSignal Labs is not a task marketplace where thousands of workers race to grab tasks before they disappear. Projects are staffed deliberately: you're matched and accepted onto a project, given training and clear guidelines, and assigned a queue of work. This means less time refreshing your dashboard and more predictability once you're on a project.
Feedback, not silent removals
Your work is quality-reviewed, and quality standards are communicated as part of each project's guidelines. If your quality scores drop, the goal is coaching and improvement first. If you're ever removed from a project, you can ask why and request a review — see Removal, disputes & appeals.
[TODO: confirm the specifics of our notice-before-removal and appeals commitments with the workforce ops team so this section can make concrete promises.]
Professional tooling
We're the makers of Label Studio, the world's most popular open source labeling platform. You'll work in Label Studio Enterprise — the same platform used by leading AI teams — not in spreadsheets or ad-hoc tools.
What we ask of you in return
High standards go both ways. We expect contributors to follow project guidelines, meet quality expectations, protect client confidentiality, and communicate honestly. See Conduct & confidentiality and Quality & reliability.