
Mateo Vargas
Senior backend engineer
11 years of experience
Reviews Python services and writes the reference solutions models are scored against.
Credexus fast-tracks software engineers in Latam, Asia, Europe and Canada into remote, high-paying AI training and evaluation work.
Independent agency. Listing a platform does not imply endorsement; trademarks belong to their owners.
Instead of guessing which of dozens of AI training platforms to try, you get a shortlist ranked by how well your skills match what they're hiring for.
We read your experience, stack and written English once — no repeated applications.
Your profile is mapped to the platforms and projects where you are most likely to be accepted.
You get briefed on what each platform's reviewers look for before you take their test.
Senior Python Engineer
São Paulo, Brazil · English C1 · 6 yrs
Outlier
Python code evaluation
Mercor
Agentic coding tasks
Alignerr
Reasoning & review
Illustrative example of a Credexus match
Three kinds of projects make up most of what our engineers do. Pick one to see a sample task.
// Prompt: dedupe a list while keeping order
return list(set(xs))
seen = set() return [x for x in xs if not (x in seen or seen.add(x))]
Rationale: B preserves insertion order in O(n); A's set() reorders items.
Compare model-written solutions, find the bugs, and explain which answer is better and why. Your rationale becomes training signal.
See rolesFrontier AI is built on expert human judgement. Yet thousands of brilliant engineers across Latin America, Asia, Europe and Canada never hear about the platforms that pay for it, or how to pass their assessments. Credexus closes that gap: one application, honest matching and a human in your corner.
AI labs need experienced engineers to write, review and rank code. There are dozens of platforms doing this work — we know which ones fit your stack, level and region.
19
AI training platforms our candidates work on
4
Regions: Latam, Asia, Europe and Canada
$0
Cost to engineers — platforms pay us
1
Application to be matched across all of them
Platforms pay us. You never pay a fee or give up a share of your earnings.
One application gets you matched across 19 platforms instead of guessing which to try.
Remote, flexible projects open to engineers across Latam, Asia, Europe and Canada.
We point you to specialist and agentic projects as your track record grows.
We share your profile with a platform only after you agree to it.
Real recruiters who answer questions about assessments, projects and pay.
A simple, human-led process designed to get strong engineers earning quickly.
Share your experience, stack, region and English level in a 3-minute application. Attach your CV — no cover letter needed.
Our team reviews your profile and maps it to the platforms and projects where you are most likely to be accepted and paid well.
We brief you on each platform's assessment, share what reviewers look for, and introduce you where we have a direct line.
Pass onboarding, pick up projects, and get paid directly by the platform. We stay in touch to help you level up to higher-paying work.
Flexible, remote contracts for engineers who care about getting the details right.
Review, debug and rank code written by large language models, and write reference solutions that teach models to code better.
Design realistic multi-file engineering tasks and grade how AI coding agents solve them in real repositories.
Compare and rate model answers to backend engineering questions, from concurrency bugs to API design.
We focus on four regions where great engineers are underrepresented on AI training platforms.
Fictional profiles that show the experience we look for across our four regions. Real engineers apply with their own background.

Senior backend engineer
11 years of experience
Reviews Python services and writes the reference solutions models are scored against.

ML evaluation engineer
8 years of experience
Scores model answers on distributed systems and marks where the reasoning fails.

Full-stack engineer
9 years of experience
Ranks generated TypeScript for correctness, clarity, and test coverage.

Staff software engineer
13 years of experience
Leads backend code reviews and steers models toward production-grade diffs.
Apply in three minutes. We will match you with the AI training platforms that fit your stack — free, always.