Python Infrastructure Engineer — LLM Training & Agent Tooling [US‐CA]

OpenTrain AI, Inc.; United States; Remote; USD 47,000 - 58,000; Part time

Job summary

OpenTrain AI, Inc. is seeking a part-time, fully remote contractor to build Python infrastructure for LLM training and agent evaluation.

You will create secure sandboxes, task frameworks, scoring pipelines, and reusable Python repositories, with a focus on developer environments that help AI researchers iterate on tasks. The role requires deep Linux and Docker knowledge, experience with FastAPI or Flask back ends, pytest-driven testing, and the ability to guide researchers within a collaborative

Qualifications

• At least five years of hands-on Python experience.
• Deep Linux and Docker knowledge; experience designing CI/CD pipelines.
• Back-end skills with FastAPI or Flask.
• Testing with pytest-driven approaches.
• Familiarity with devcontainers and Makefiles.

Responsibilities

• Build and maintain infrastructure behind LLM training and agent evaluation.
• Create reusable Python repositories and developer environments.
• Develop secure sandboxes for agent tasks and scoring pipelines.
• Design task frameworks and automated scoring pipelines.
• Guide AI researchers through infrastructure and tooling.

Skills

Python

Collaboration

English communication

Hands-on experience

Tools

Docker

Linux

CI/CD pipelines

FastAPI

Flask

pytest

devcontainers

Makefiles

GitHub Actions

Job description

Build Python infrastructure for LLM training and agent evaluation, including secure sandboxes, task frameworks, and scoring pipelines. This remote contract offers 20+ hours weekly and $34 to $42 per hour in the US or Canada.

Remote Hourly · $37/hr

$37/hr

Compensation

Contract, part-time

Engagement

Remote

Location

Jul 25, 2025

Posted

Open to applicants in

United States Canada

The Work

You will build and maintain the infrastructure behind LLM training and agent evaluation. The work focuses on reusable repositories, secure sandboxes, task frameworks, scoring pipelines, and developer environments that help AI researchers iterate on agent tasks.

You will also help researchers use these tools and contribute to a collaborative engineering workflow.

• Build secure sandboxes for agent tasks.
• Create task frameworks and automated scoring pipelines.
• Develop reusable Python repositories and developer environments.
• Design CI/CD pipelines, with GitHub Actions preferred.
• Build or support FastAPI and Flask back ends.
• Use pytest-driven testing to check code and evaluation workflows.
• Use devcontainers, Makefiles, and other modern development tools.
• Guide AI researchers through the infrastructure and tooling.

What It Pays And Takes

This is a part-time contract role for candidates in the United States or Canada. The listed experience level is intermediate, and compensation is tiered by experience.

• Pay: $34 per hour for junior contributors, $37 per hour for middle-level contributors, and $42 per hour for senior contributors, in USD.
• Time: 20 or more hours per week.
• Location: Fully remote for candidates in the United States or Canada.
• Language: Clear English communication is required.
• Experience: At least five years of hands‑on Python experience.
• Technical skills: Deep Linux and Docker knowledge, plus experience designing CI/CD pipelines.
• Back-end skills: Familiarity with FastAPI or Flask is required.
• Testing: Experience with pytest-driven testing is required.
• Tooling: Familiarity with devcontainers and Makefiles is required.
• Working style: A collaborative mindset and the ability to guide AI researchers are required.
• Bonus: Prior work on OpenTrain coding tasks is a strong plus.
• Selection requirement: Short‑listed candidates complete a timed HackerRank assessment and a platform coding test before recruiter interviews.

About AI Training Work

AI training work is the human side of building artificial intelligence. People create code, examples, and evaluations that help AI systems learn how to perform tasks, and experienced engineers are paid to make the tools and tests behind that work reliable. OpenTrain helps people find and build careers in this field.