Senior Backend Engineer

What you'll do

You build the exam. Each task is a reinforcement learning environment in which an AI model has to design, deploy, troubleshoot, secure, scale or recover a piece of production-grade cloud infrastructure, and your job is to make that environment realistic enough that passing it means something.

A complete task has four parts:

• A scenario drawn from the messy middle of infrastructure work: distributed systems, networking, IAM, message queues, durable storage, observability, rolling deployments or disaster recovery
• A reproducible environment that stands up the same way every time
• Deterministic validation tests plus a golden reference solution that passes them
• Intentionally defective variants, broken on purpose, so the model has to diagnose rather than pattern-match

You also document the architecture, edge cases and operational flow of each environment, and iterate on specifications and acceptance criteria with the technical leads.

Who fits

A backend engineer who has been paged. The ad asks for:

• Strong command of a backend language: C++, Python, Rust, Go, Java or JavaScript
• Practical DevOps experience with cloud infrastructure, CI/CD pipelines and automation tooling
• Experience architecting, scaling and securing distributed systems in production
• Real understanding of networking, IAM, queues, durable storage and disaster recovery

No prior AI experience is needed; the ad says domain knowledge is what matters. No minimum years are stated, but the qualifications describe senior production experience.

What it pays

$50–100/hour is the published band, but compensation is output-based: you are paid per task that meets the project specification, and micro1 notes that time per task varies with experience and workflow. A minimum number of tasks per week applies, though the ad does not give the number.

In practice the hourly figure is a target. Designing a deterministic test for a failure-recovery scenario can take far longer than expected the first time, and time on a task that is not accepted is not paid. Ask during screening what a typical task takes and what the weekly minimum is.

Commitment and process

• About 20 hours a week, contractor, fully remote
• Apply and answer screening questions, then an AI interview of roughly 30 minutes reviewed by recruiters, then a hiring manager review
• micro1 says it usually fills roles within 48 hours and expects first tasks within 24–48 hours of onboarding

Worth knowing

Good:

• Genuinely senior engineering work: building reliable test harnesses for infrastructure failure is hard and interesting
• Seven languages accepted, so you are not locked to one stack
• Defined weekly commitment of about 20 hours
• Short, clearly described hiring process
• Ten openings

Less good:

• Per-task pay, so the hourly band is not guaranteed and rejected work is unpaid
• A minimum weekly task count with no number published
• Fast start expectations: be ready to begin within two days of onboarding
• Listed for about two months, so some openings may already be filled
• Contractor engagement: no benefits and your own tax to manage

About this listing

Posted by micro1 on its own jobs portal about two months before we read it on 24 September 2026. The $50–100/hour band, ten openings, 20-hour weekly commitment and per-task compensation terms are its published text. No country restriction is stated and the customer is not named. See micro1.