Fully remote tasks
Review batches, rate responses, and leave notes from any timezone — all through our contributor portal.
Human-in-the-loop · Language QA
Turn language expertise into paid AI work — from anywhere.
Cognitivus Labs recruits reviewers who evaluate LLM outputs, refine prompts, and score response quality for production AI systems. Structured briefs, async tasks, and pay in PHP.
Start your application
Typical rate
PHP 180 – 450 / hr
Task approved · Paid
24h
Typical review turnaround
Why Cognitivus
Language work that fits around your life — not the other way around.
Async scheduling
Claim tasks when you're free. No fixed shifts, no camera-on meetings.
Short skills check
One language assessment, a sample review task, then paid project queues matched to your profile.
PHP weekly payouts
Approved tasks are tallied every week. Higher accuracy unlocks bonus rates and priority queues.
Open position
AI Language Reviewer
Production LLMs still need human judgment — to catch awkward phrasing, cultural mismatches, and factual drift before outputs reach users.
We're building a reviewer pool across 50+ languages. English fluency is required; a second language opens more project types.
You'll do
- Score model replies against rubrics for clarity, tone, and factual accuracy
- Edit or rewrite weak responses and explain what failed
- Compare answer pairs, rank outputs, and flag policy or safety issues
We need
- Native-level feel for at least one language you apply with
- Ability to explain edits clearly — not just “sounds wrong”
- Steady output and consistent scores across review batches
How it works
From application to first paid task in days, not weeks.
01
Apply with your languages
Tell us your fluency, availability, and background. We reply with next steps — usually within 48 hours.
.avif)
02
Matched to live queues
Work on active batches: pairwise ranking, rewrite tasks, prompt fixes, and domain-specific QA (legal, medical, finance, and general).
03
Track hours and payouts
See approved tasks, hourly totals, and weekly PHP transfers in your contributor dashboard.
Get started
Three steps to your first task
-
1
Submit the form
Name, email, primary language, and weekly hours — under two minutes.
-
2
Pass the review test
A 20–30 minute exercise: rate sample LLM outputs and fix one short response.
-
3
Get task access
Unlock paid queues, submit work, and receive weekly PHP payouts for approved batches.
FAQ
Questions before you apply?
Tasks, pay, and eligibility — straight answers before you apply.
What kind of tasks will I do?
Most work is linguistic QA: ranking two answers, rewriting a weak reply, scoring tone and accuracy, or fixing a prompt. Some batches add light fact-checking. No engineering background needed.
What projects are available?
Queues rotate by client need — common types include RLHF preference labeling, SFT output review, RAG answer checks, and red-team style safety evals.
Who is this best for?
Writers, editors, translators, teachers, researchers, and bilingual professionals who notice when text sounds off — and can say why.
What's the work arrangement?
Independent contractor, fully remote. You choose your hours and volume; we send weekly PHP payments for approved work.
How does payment work?
Rates start at PHP 180/hr depending on task type and language. Approved hours are paid out weekly to your nominated account. Quality bonuses apply on high-accuracy streaks.
Open applications
Ready to contribute?
Two-minute form, one review exercise, then paid task queues. No recruiter calls or multi-round interviews.