Job description

ABOUT MECKA AI Mecka AI is building the data infrastructure layer for robotics and embodied AI. We design and operate global systems for data capture, data labeling, and hardware-enabled workflows used by leading AI labs and robotics companies to train and validate humanoid and embodied AI systems. We work closely with frontier robotics teams to bridge real-world data, simulation, learning-based systems, and deployed hardware. The Role The Research Engineer, RL Env role at Mecka AI involves building reinforcement learning environments that help researchers train models and understand their capabilities. Across Mecka's Labs team, you'll translate real tasks into computational problems, implement environments for model training and test whether measured progress reflects useful behavior. This is a hands-on research engineering role for someone who knows how to construct reinforcement learning environments. You'll write working software, investigate failures and develop methods with product colleagues, domain experts and engineers. WHAT YOU WILL BE DOING: - Build environments: Define tasks, observations, actions and state transitions. Implement reset behavior, termination conditions and measurable outcomes in environments agents can interact with. - Develop rewards and scoring: Translate task objectives into feedback and evaluation criteria. Test whether agents can exploit scoring rules without completing the intended task. - Run learning experiments: Implement baseline agents, train and compare policies, and design controlled experiments that isolate the effects of data, methods and environment changes. - Make evaluation reliable: Separate training and held-out tasks, check for leakage, version experiments and repeat runs. Report uncertainty and performance across conditions alongside aggregate scores. - Investigate failures: Inspect trajectories and learning behavior to distinguish policy limitations from data, reward or environment problems. Use findings to prioritize the next experiment. - Build with the team: Work with domain experts to validate task assumptions and with engineers to turn research prototypes into reusable environments, evaluation tools and documented methods. WHAT YOU BRING: - RL environment expertise: You have hands-on experience formulating problems, constructing environments, training agents and critically assessing results. - Strong programming and software debugging skills; able to build and test research systems that other people can run and extend. - Sound experimental design and statistical reasoning, including controlled comparisons, evaluation splits, variability and the limits of benchmark results. - Ability to reason about environment dynamics, reward design and agent behavior, and trace unexpected results to concrete causes. - Independent research judgment and clear communication; learn unfamiliar domains, work with specialists and explain assumptions, tradeoffs and findings. EVEN BETTER IF YOU HAVE: - Experience building interactive environments, simulators or benchmarks used by other researchers. - Work on agent evaluation, reward design, imitation learning or learning from real-world data. - Research artifacts with reproducible experiments, useful baselines and evidence of investigating failures beyond headline scores. If you don't check every box above but believe you can do the job, we encourage you to apply — we're looking for capability and trajectory, not a perfect checklist match. We strive to create an inclusive work environment for all; we welcome applicants from all backgrounds and are committed to providing equal opportunities. Should you require any accommodations during the recruitment process, we will partner with you to meet your needs. We may use AI-assisted tools to help evaluate applications against the requirements of this role. These tools support our recruiting team in identifying qualified candidates but do not replace human judgment or limit which applications may be reviewed or considered.

Requirements

Must-haves

Build environments: Define tasks, observations, actions and state transitions. Implement reset behavior, termination conditions and measurable outcomes in environments agents can interact with.
Develop rewards and scoring: Translate task objectives into feedback and evaluation criteria. Test whether agents can exploit scoring rules without completing the intended task.
Run learning experiments: Implement baseline agents, train and compare policies, and design controlled experiments that isolate the effects of data, methods and environment changes.
Make evaluation reliable: Separate training and held-out tasks, check for leakage, version experiments and repeat runs. Report uncertainty and performance across conditions alongside aggregate scores.
Investigate failures: Inspect trajectories and learning behavior to distinguish policy limitations from data, reward or environment problems. Use findings to prioritize the next experiment.
Build with the team: Work with domain experts to validate task assumptions and with engineers to turn research prototypes into reusable environments, evaluation tools and documented methods.
RL environment expertise: You have hands-on experience formulating problems, constructing environments, training agents and critically assessing results.
Strong programming and software debugging skills; able to build and test research systems that other people can run and extend.
Sound experimental design and statistical reasoning, including controlled comparisons, evaluation splits, variability and the limits of benchmark results.
Ability to reason about environment dynamics, reward design and agent behavior, and trace unexpected results to concrete causes.
Independent research judgment and clear communication; learn unfamiliar domains, work with specialists and explain assumptions, tradeoffs and findings.
Experience building interactive environments, simulators or benchmarks used by other researchers.
Work on agent evaluation, reward design, imitation learning or learning from real-world data.
Research artifacts with reproducible experiments, useful baselines and evidence of investigating failures beyond headline scores.
If you don't check every box above but believe you can do the job, we encourage you to apply — we're looking for capability and trajectory, not a perfect checklist match.

Nice-to-haves

None flagged.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

Platform: ashby
Effort: quick