Senior Software Engineer (RL Environments, $100–150/hr)

What you'll do

Each unit of work is a small, self-contained software engineering problem packaged so an AI model can be dropped into it and scored. You pick a realistic task (a defect to fix, a feature to add, a module to refactor, a hot path to speed up), build a reproducible environment around it, and write the golden reference solution the model's attempt is measured against.

The ad frames this as reinforcement learning environment work, which sets two requirements. The environment has to behave the same way every time it is rebuilt, and the reference has to be the answer a strong engineer would actually ship, because a sloppy reference teaches the model the wrong thing. Around that core you document your reasoning and solution approach, and review code and submissions from other contributors for accuracy.

Languages named are Python3, Java, Rust, Go, C++ and TypeScript; you need depth in at least one.

Who fits

• Significant hands-on work in at least one of Python3, Java, Rust, Go, C++ or TypeScript
• Solid algorithms, data structures and engineering fundamentals
• A record of debugging hard problems in complex systems and landing real performance wins
• Experience refactoring large or legacy codebases
• Features shipped end to end in cross-functional teams
• Clear technical writing, since your explanation travels with the task

No prior AI experience is required, and no degree or minimum years are stated.

What it pays

$100–150/hour is the published band, but compensation is output-based: you are paid per task that meets the project specifications, with a weekly minimum number of submissions (the ad does not give the number). Time per task varies with your experience and workflow, so your effective rate is really your acceptance rate multiplied by your speed. The expected load is about 15 hours a week.

How it differs from micro1's other Senior Software Engineer listings

micro1 runs several postings under the same title. This one carries the highest band of the RL environment postings we have read, and its text is word for word the same as a $50–100/hour sibling, so the project looks the same and only the band differs. A separate MCP-focused posting at $80–120/hour adds Model Context Protocol tool use to the tasks. The Coding Tasks and Verifiers listing at $30–100 centres on writing verifiers, and the $245–280 listing is code evaluation with a big-tech pedigree requirement.

Worth knowing

Good:

• The top band among micro1's RL environment postings, with a $100 floor
• 100 openings, flagged high demand, posted within the last week
• About 15 hours a week, workable next to a day job
• No AI background or employer pedigree required
• Real engineering work: debugging, refactoring and performance, not labelling

Less good:

• Per-task pay: rejected tasks are unpaid and no per-task fee is published
• A weekly minimum applies but is not quantified
• Making an environment truly reproducible often takes longer than solving the problem
• Fast start expected: roles fill in about 48 hours and first tasks are due 24–48 hours after onboarding
• Contractor engagement: no benefits, your own tax to manage

About this listing

Posted by micro1 on its own jobs portal and read on 24 September 2026, five days after it went up. The $100–150/hour band, the 100 openings, the 15-hour weekly guide and the per-task compensation terms are its published text. Location is Remote with no country list, and the customer is not named. Payment schedule and method are not published on the listing. See micro1.