Is AfterQuery Legit? Pay, Payment Schedule & How the Platform Works
AfterQuery Experts review: who they are, how much they pay ($35–$250/hr depending on tier), when you get paid, and what the project-based work feels like day to day.
AfterQuery is one of the newer platforms for AI training work, and it pays at the higher end of the spectrum. You'll see doctors, lawyers, PhD researchers, and senior engineers working alongside generalist annotators. That mix raises a fair question: is this a real company with real money behind it, or another platform overselling premium pay that never materializes?
It's real, it's well-funded, and it pays. But the work doesn't arrive like a steady job, it comes in waves, and that's the main thing that catches people off guard. Here's what AfterQuery is, what it pays, and what to expect once you're in.
Is AfterQuery Legit?
Yes. AfterQuery is a San Francisco-based applied research lab that builds training data and reinforcement learning environments for frontier AI labs. It went through Y Combinator's Winter 2025 batch and closed a $30M Series A in April 2026 at a $300M valuation, led by Altos Ventures and The Raine Group, with Y Combinator and BoxGroup also participating. The company has reported surpassing $100M in annual revenue run rate. Founders Spencer Mateega (CEO), Danny Tang, and Carlos Georgescu are publicly listed and verifiable on LinkedIn.
YC W25, $30M Series A (April 2026) at a $300M valuation. Clients are frontier AI labs, not anonymous "AI startups."
Generalist reviewer roles cluster around $35–$75/hr, credentialed specialists (PhDs, doctors, domain experts) see $100–$250/hr.
You apply once as an "Expert." Once accepted and onboarded, you get access to live projects and direct invitations, not a fresh application per gig.
Fixed schedule, not "whenever." Details below.
How AfterQuery Works
AfterQuery isn't building a chatbot. It's building the data that teaches frontier models how to reason through real expert work: detailed chain-of-thought prompt-response pairs for supervised fine-tuning, "agent environments" where a model has to chain actions, call APIs, run code, and recover from its own mistakes, and computer-use trajectories that teach a model to operate software the way a person does.
That requires actual domain professionals, not just generalist labelers. That's the "Experts" network: investment bankers, doctors, PhD researchers, software engineers, and generalist workers, all recruited to do graded, instructed tasks inside specific projects.
Apply once
One application gets your profile reviewed. No reapplying for each new project once you're in the network.
Get onboarded
Once your profile is approved, you unlock the dashboard. You'll see live generalist projects directly, plus invitations to domain-specific projects based on your background.
Work the project
Each project posts its own pay structure in the instructions. Per-task projects need your submission to pass review before it counts toward payment, hourly projects pay for logged time. Either way, you know what's expected before you start.
Get paid on Friday
Pay periods run Monday through Sunday Pacific Time. Whatever you completed and got approved in that window pays out the following Friday, by 11:59 p.m. PT.
Pay Rates & When You Get Paid
The pay spread is wide because the work is varied. A generalist annotator and a PhD physicist aren't doing the same task, so they don't get the same rate.
| Tier | Typical Rate | Who Qualifies |
|---|---|---|
| Generalist | $35–$75/hr | Strong written English, careful reviewer-style work, no specific credential required |
| Technical specialist | $75–$150/hr | Software engineers, data scientists, niche technical domains (e.g. legacy systems, EDA) |
| Credentialed expert | $150–$250/hr | Doctors, lawyers, PhD researchers, finance professionals with verifiable credentials |
Based on listings in our dataset. Each project sets and discloses its own rate up front.
Pay period: Monday through Sunday, Pacific Time. Whatever you finish and get approved inside that window counts for that period.
Payday: Every Friday, by 11:59 p.m. PT, for the period that just closed. It's the same day every week, no surprises.
Per-task vs. hourly: Each project's instructions state which structure it uses. Per-task work has to pass review to qualify for payment, and if a task is rejected, you're told why.
The payment guarantee: AfterQuery says it will never train on data you haven't been paid for, or that isn't already queued for payment. If a task gets rejected, it doesn't quietly get handed to clients anyway, it's excluded.
What the Work Feels Like
Here's the part people need to know going in: AfterQuery's projects don't form a steady daily queue. Work arrives in clusters. A project goes live, there's a burst of tasks to finish in a tight window, and then things can go quiet for a while as the next project gets prepared.
Because task volume is split across the whole expert pool, popular projects can fill up fast. If you're not online when a batch opens, the tasks can disappear before you get to them. Instructions can also shift mid-project as the client refines what they're actually testing, which is normal for frontier-lab work where the target behavior is still being defined.
Set Your Expectations Right
Treat AfterQuery like project-based contract work, not a part-time job with guaranteed hours. The pay ceiling is genuinely high, especially for specialists, but volume varies week to week. It's the same trade-off you see on other expert-network platforms: high rates in exchange for irregular availability, not a steady stream of tasks.
Pros & Cons
👍 The Good
- ✓ Real backing: $30M Series A at a $300M valuation, backed by Altos Ventures and The Raine Group. Real clients, not a side hustle dressed up as a company.
- ✓ Strong specialist pay: $150–$250/hr for credentialed experts beats most other platforms.
- ✓ Predictable paydays: Fixed Monday–Sunday PT pay periods, paid every Friday. No guessing when money arrives.
- ✓ One application: Get into the Expert network once, not per project.
👎 The Bad
- × Work comes in waves, not steady streams: You might earn well one week and have nothing the next.
- × Popular projects fill up fast: Checking in late to a fresh batch can mean there's nothing left.
- × Per-task rejections don't pay: Read the instructions carefully before you submit.
- × Newer platform: Less public track record compared to established names like Outlier or Mercor.
Should You Apply?
YES, IF:
You have a domain credential (medical, legal, finance, PhD research) or solid technical depth, and you're comfortable working in focused bursts when projects open.
PROBABLY NOT IF:
You need predictable hours every week. AfterQuery pays reliably, but the work itself is project-based and sporadic.
Related guides
Platform resources: Browse open AfterQuery roles · AfterQuery platform profile · Alternatives to AfterQuery
Other platform reviews: Mercor · Alignerr · Micro1 · Full platform comparison
Context: What is fine-tuning? · How to become an AI trainer
Pietro R.
MSc Human-Computer Interaction | Founder & Product Owner
Pietro is the founder and technical lead of aitrainer.work. He builds and maintains the platform's data pipeline, certification infrastructure, and editorial standards.