Member of Technical Staff, Research Engineering

What you'll do

This salaried micro1 role is about the machinery of reinforcement learning rather than grading its output. You architect self-contained RL environments for complex real-world tasks, complete with reward functions, verifiers and evaluation logic; scale episode pipelines and multi-component training processes for reproducible experiments; and build synthetic data generation that speeds up training without lowering quality.

The evaluation thread runs throughout: AI-driven grading and validation loops, benchmarking frameworks for capability and robustness, and publishing evaluations on internal and external platforms, including micro1's own benchmarks. You also fine-tune open-source models on internally generated data.

Who fits

• Deep RL experience, including environment design and training dynamics
• A record of building and scaling RL systems or experimentation frameworks
• Automation and synthetic data pipeline experience
• Familiarity with automated evaluation, model validation and QA workflows
• Hands-on fine-tuning and evaluation of open-source models
• Concise technical writing, and comfort in a fast research environment

Preferred: published benchmarks or research artifacts, familiarity with evaluation ecosystems, and scalable RL infrastructure.

What it pays

The compensation notice gives a national base salary range of $200,000–300,000 USD, with equity eligibility and possible performance bonuses. Benefits include up to 100% reimbursement of health-insurance premiums, paid time off and a 401(k) with company match.

The listing's pay widgets disagree with that: the card shows $400,000–800,000/year, which may include equity (the ad does not say), and the detail panel shows an hourly figure that is clearly not a salary. We list the base range from the written notice.

Worth knowing

Good:

• Salaried with equity and benefits
• Upper end of the base range is higher than micro1's Coding Research role ($200,000–260,000)
• Real ownership of RL environment design, not just data production

Less good:

• One opening, for a specialist profile
• Conflicting pay figures on the listing; confirm the total package directly
• No hiring countries stated, though the 401(k) points to US employment
• AI tools are used in candidate screening, per the ad

About this listing

Posted by micro1 on its own jobs portal 11 days before we read it on 24 September 2026. The base salary range, benefits and requirements come from that listing's text; the ad includes a disclaimer that it is informational and not a binding offer. Visa sponsorship is not addressed. See micro1.