Free · built from your skills
Learn AI training with a lab built from your skills
Learn AI training by doing the work an AI trainer does: compare model responses, grade them with rubrics, write your own rubrics and design a task a model gets wrong. Your lab is generated from your résumé, so every task is in your own field, and every answer gets feedback.
Ready in a minute or two. Work at your own pace; your progress is saved.
Made for your skillset, not a generic course
The lab reads your role and skills, then writes its questions, response pairs and tasks around them, so people in different fields get different labs. For example:
Nursing
You rate two answers about a medication question and say which a charge nurse would trust.
Accounting
You grade a response on a month-end reconciliation against a weighted rubric.
Software
You write a rubric that separates a correct SQL fix from a confident wrong one.
Law
You design a task from your own practice area that a model gets partly wrong.
What you'll learn: seven modules
The kinds of work AI training platforms test before your first paid task, in the order projects use them.
Module 1
How human data trains models
Demonstrations, preference pairs and rubrics, and how your labels become a model's training signal.
Module 2
Reading project guidelines
How guidelines decide what “better” means, and how to apply them under pressure.
Module 3
Side-by-side preference
Compare two responses, rate which is better and by how much, and tag what went wrong.
Module 4
Grading with a rubric
Apply a weighted rubric to a response, criterion by criterion.
Module 5
Justification writing
Write the justification a reviewer could check.
Module 6
Rubric authoring
Write a rubric that tells a strong response from a flawed one.
Module 7
Task design capstone
Design a task from your own field that a model gets partly wrong.
What a task looks like
A preference pair from module 3, for a software candidate. Read both responses, then decide which is better and by how much.
WHERE id NOT IN (SELECT user_id FROM orders) returns no rows, but some users have no orders. Why?Response A
One of the subquery's user_id values is NULL. NOT IN compares against every value, and a comparison with NULL is unknown, so no row qualifies. Use NOT EXISTS, or filter the NULLs out of the subquery.
Response B
This is a known bug in PostgreSQL 15 that was fixed in version 16. Upgrade your database and the query will return the users with no orders.
A is much better.
B invents a bug and a fix: a confident fabrication that would send the user after an upgrade that changes nothing. In the lab, you would also tag B for hallucination and write one or two sentences a reviewer could check.
How the lab works
- 01Add your résuméSign up free and upload it. We read your role and skills; that is what the lab is built from.
- 02Your lab buildsAbout a minute or two. The first module is ready first, so you can start while the rest builds.
- 03Work through seven modulesFrom how human data trains models to designing your own task. Each module opens when you finish the one before it, passed or not.
- 04Learn from every answerEach answer is scored the way a project reviews it, against a reference answer, with the reasoning shown.
How to become an AI trainer, starting with the lab
AI training platforms hire an AI trainer for expertise you already have, and test you on exactly this work before your first paid task. Practise in the lab until the task types feel routine. When you are ready, browse open AI training jobs in your field, or compare the platforms that hire. New to the field? Start with how AI models are trained.
FAQ
Learning AI training: common questions
How do I learn AI training?
By doing the tasks AI labs pay for: comparing two model responses, grading a response against a rubric, writing a rubric, and designing a task a model gets wrong. Swooply's free lab gives you those tasks in your own field, built from your résumé, with feedback on every answer.
How do I become an AI trainer?
Most AI training platforms hire for expertise you already have (medicine, law, accounting, software, languages, science) and test you before your first task. Learn the workflow, practise on realistic tasks in your field, then apply to roles that match your background.
What qualifications do you need to be an AI trainer?
Usually expertise you can show: a degree, professional experience or fluency in a language. Platforms then test you on the work itself (rating responses, grading against a rubric, writing clear explanations) before your first task. The qualification that matters most is being right about your own field and saying why.
Can you become an AI trainer with no experience?
With no AI experience, yes; with no expertise, rarely. Platforms hire nurses, accountants, lawyers and engineers for what they already know, and teach or test the AI side. Practising the task types first, in your own field, is the part you can do before you apply.
Is it hard to become an AI trainer?
The work is not technical, but the qualification tests are strict: careless ratings, vague explanations and missed errors fail. Applicants who have practised comparing responses and grading against rubrics know what the tests look for.
Is the lab free?
Yes. The lab is free for candidates. You sign up with your résumé and the lab is built from it.
Do I need to know how to code?
No. Most AI training work is judgment, not programming. The lab is built from your own field and skills, so the tasks are about the work you already know.
How long does it take?
The lab is ready in a minute or two, and you can start the first module while the rest builds. You work at your own pace and your progress is saved.
What happens if I don't pass a module?
You keep going. The next module opens when you finish the one before it, whether or not it passed, and every answer comes with an explanation of what it was testing.