Build MCP servers and integrate real applications into an RL training sandbox using Python, Docker, and Linux. Contract, remote role: 20 hours/week, $40–$50/hr for engineers comfortable with back-end APIs, debugging, and deployment pipelines.
Coding & Software
100% Remote Hourly · $40–$50/hr
$40–$50/hr
Compensation
Worldwide
Eligibility
Entry
Experience
Jul 21, 2026
Posted
Open worldwide
Interested in this role?
Create a free OpenTrain account and apply in minutes.
OpenTrain is the #1 platform for people who build careers in AI training and data labeling. We connect contributors with hands-on engineering and annotation work, help them track projects in one place, and make it simple to grow a durable freelance career.
Creating an OpenTrain account is free — it lets you apply, build a unified profile of AI training experience, and manage contracts and ongoing projects.
Why AI training work matters
AI training is the human side of building modern intelligence: engineers and annotators create, evaluate, and refine the examples that teach models how to behave. This role puts you on the engineering side of that process, building reliable infrastructure for reinforcement-learning experiments.
Work in this space is frequently remote and flexible, and contributors directly influence how state-of-the-art systems are trained and validated.
100% remote: work from anywhere with a reliable internet connection.
Flexible, part-time-friendly assignments that fit around other commitments.
The role
OpenTrain AI is recruiting a Software Engineer to build and ship applications for an RL (reinforcement-learning) platform. You will develop MCP servers, integrate real web or desktop applications into an RL Studio sandbox, and ensure apps are stable, testable, and deployment-ready.
This is a contract, part-time engagement (20 hours per week) and is fully remote.
Position type: Contractor, part-time (20+ hours/week).
Location: Fully remote, worldwide; English required.
What you'll do
Build MCP servers in Python using FastMCP and similar frameworks.
Integrate web or desktop applications into an RL Studio sandbox environment.
Work with APIs, Docker containers, Linux environments, and service configuration.
Implement populate/snapshot hooks to manage and restore app state.
Test applications locally and debug build, runtime, and permission issues.
Ensure applications meet platform validation and deployment requirements.
Requirements
Strong Python development skills.
Experience building and integrating APIs and backend systems.
Familiarity with Docker, Linux, and local testing workflows.
Comfort working with configuration files and deployment pipelines.
Proven ability to debug build-time, runtime, and permission-related issues.
Attention to technical detail and strong problem-solving skills.
Helpful background
Experience with FastMCP or similar MCP frameworks.
Prior work integrating third-party applications or developer tools into sandboxes.
Familiarity with web applications, databases, and containerized deployment environments.
Pay, schedule, and logistics
Compensation is $40–$50 per hour (USD), paid per hour and based on experience and qualifications.
Schedule: approximately 20 hours per week with an ongoing commitment. This contract work is fully remote and open worldwide.
How it works / Next steps
If this role fits your skills and availability, create a free OpenTrain account and apply. Your OpenTrain profile helps showcase relevant project experience and makes it easy to manage contract work.
OpenTrain AI will review applicants and contact qualified candidates for technical evaluation and onboarding details.
Join OpenTrain AI to build the Python infrastructure that powers LLM training and agent evaluation. Fully remote for Asia‑Low region, 20+ hrs/week contract work with tiered hourly pay; requires 5+ years Python, Docker, CI/CD, FastAPI/Flask, and Linux expertise.
OpenTrain AI seeks an experienced Docker-focused code reviewer to validate annotator reviews of AI-generated Dockerfiles and orchestration snippets; $24/hr, 20+ hrs/week, remote contractor role. Run builds in sandboxed VMs, flag issues, and provide concise remediation notes to meet a strict quality
Build and own the infrastructure that powers LLM training and agent evaluation at OpenTrain AI; remote for US & Canada, 20+ hours/week, contractor roles with hourly pay tiers (Junior $34 / Middle $37 / Senior $42). Shortlisted candidates complete a timed HackerRank and platform coding test.
