Software Engineer for AI Training Environments

About OpenTrain

OpenTrain is the #1 platform for people building careers in AI training and data labeling. We help contributors discover projects, build a unified portfolio, and grow durable freelance careers teaching AI systems.

OpenTrain AI is the hiring and contracting organization for this role. This position is a contractor engagement focused on creating high-quality software assets used to evaluate AI systems.

About AI training and this kind of work

AI training (also called data labeling or human feedback work) is the human side of how modern models learn. Contributors create the examples, tests, and evaluations that shape model behavior.

This role sits at the intersection of software engineering and AI evaluation: you build deterministic, reproducible programming tasks and scoring logic so models can be measured on realistic development scenarios.

The role

Title: Software Engineer for AI Training (contractor, part-time). You will design, implement, and maintain reinforcement-learning environments and software tasks used to evaluate model performance.

Commitment: less than 20 hours per week. Location: fully remote, worldwide. Language: English. Compensation: PAY_PER_HOUR with an hourly range of $50–$150 and a top rate of $150/hr.

• Employment type: Contractor, Part-time.
• Data type: computer code / programming; label types include fine-tuning, evaluation rating, and programming/coding evaluation.

What you'll do

You will produce reproducible, high-signal software tasks and scoring harnesses that let we evaluate model tool use and engineering ability in realistic scenarios. Work emphasizes deterministic verification and maintainability.

• Build and refine reinforcement-learning environments for AI evaluation workflows.
• Create software engineering tasks that require models to reason over real Model Context Protocol (MCP) server information.
• Write deterministic checks and golden reference solutions for reproducible scoring.
• Improve code quality via debugging, feature implementation, refactoring, and performance tuning.
• Support evaluation of model behavior in realistic, high-signal programming scenarios.

Requirements

You must be able to deliver maintainable software and clear written explanations of technical decisions. This is an entry-level friendly posting but requires solid engineering fundamentals.

• Strong proficiency in at least one: C++, Python, Java, Go, TypeScript, or Rust.
• Deep understanding of algorithms, data structures, debugging, and performance optimization.
• Experience debugging complex software issues and delivering maintainable fixes.
• Comfort with feature development, codebase refactoring, and scalable engineering practices.
• Ability to design reproducible engineering tasks and deterministic verification.
• Strong written and verbal communication skills.

Helpful background (nice-to-have)

The following experience will make you more effective, but none are strict requirements unless stated above.

• Experience with large-scale distributed codebases.
• Familiarity with modern AI or machine-learning systems and evaluation practices.
• Experience participating in rigorous code reviews and applying software best practices.

How it works and how to apply

OpenTrain makes it simple to apply and get started: create a free OpenTrain account, complete a profile listing your languages and technical skills, and submit your application for this contractor role. OpenTrain AI manages contracting and payment.

If selected you'll work remotely on short, focused tasks and deliverables. Pay is hourly (PAY_PER_HOUR) within the stated $50–$150/hr range; top rate available is $150/hr. Time commitment is flexible but expected to stay under 20 hours/week.

• Languages required: English.
• Worldwide applicants welcome.
• Typical deliverables: environment code, deterministic tests, golden references, and documentation.