- Work from
- Worldwide
- Contract
- Full-time
- Salary
- $35 – $70 / hour
- Posted
- 18 November 2026
- Apply by
- 17 January 2027
You will help build training environments and evaluate agent behavior, contributing to AI systems that interact with software and real-world workflows.
About the role
Join a paid talent pool supporting AI projects focused on agents and reinforcement learning.
Responsibilities
- Build reinforcement learning environments using Python and Docker so agents can act and receive feedback.
- Model business workflows, including ERP processes, CRM updates, and spreadsheet operations, as trainable scenarios.
- Write tasks, connect tools and REST APIs, and create systems to verify agent actions.
- Create and grade agent demonstrations, including screen recordings of task execution.
- Collect execution traces, evaluate agent trajectories, and debug failures.
Requirements
- Zero to five years of experience as a software engineer, QA automation engineer, ML or RL engineer, operations analyst, or business software power user.
- Experience writing Python or building automations or reports in Excel, Salesforce, or an ERP system.
- Experience testing, scripting, or automating multi-step software tasks, such as test cases, API calls, or UI steps.
- Familiarity with business workflows and operations, or strong software fundamentals.
- Experience with Docker, REST APIs, Gymnasium or OpenAI Gym-style environments, pytest, Playwright, or Selenium is useful.
- Experience documenting business workflows or recording screen-based tasks is useful.
Salary
Compensation is $35 to $70 per hour, depending on the project, field, and experience level.
Compensation & Benefits
Pay is based on hours worked.
Location
Remote. Projects typically last from one week to several months.
B
A
F
M
T
P
Discussion
No comments yet. Ask a question or share a tip about this role.