Software Engineering AI Task Author

About OpenTrain

OpenTrain AI is the hiring and contracting organization for this role and the #1 platform for finding and building careers in AI training and data labeling. OpenTrain helps contributors discover projects, build a professional profile, and grow a long-term portfolio in this fast-moving field.

About AI Training Work

AI training is the human work behind modern artificial intelligence. Contributors create examples, evaluate model behavior, and provide precise feedback that helps AI systems solve real-world problems more reliably. This role applies hands-on software engineering judgment to the development and evaluation of advanced AI agents.

• Fully remote work with flexible scheduling
• Part-time, task-based contractor opportunity
• Directly contribute to cutting-edge AI development

The Role

OpenTrain AI is seeking a Software Engineering AI Task Author to create expert-level tasks based on realistic enterprise engineering work. You will transform integration failures, logs, codebase bugs, configuration problems, and system states into practical challenges that test an AI agent's ability to diagnose and resolve software issues.

This is a freelance, task-based role for engineers who can translate production-oriented experience into precise prompts, realistic task environments, and checkable evaluation criteria. The role is listed as entry level, while the work requires current hands-on software engineering experience in industry-level environments.

• Contractor, part-time engagement
• Expected commitment: 20+ hours per week
• Pay range: $70–$120 USD per hour
• Work available worldwide
• Primary working language: English

What You'll Do

You will author realistic engineering task prompts and build the surrounding environments needed to evaluate AI-generated solutions. Each task should reflect practical engineering standards and include a clear, verifiable definition of success.

You will also solve the tasks yourself, validate the resulting fixes, calibrate difficulty, and improve AI-drafted materials when needed.

• Create prompts covering failing integrations, log-based diagnosis, source-code debugging, configuration changes, and integration fixes
• Write scoring rubrics that define correct diagnoses, acceptable fixes, and verification methods
• Build task environments containing representative codebases, logs, alerts, and system states
• Implement fixes and confirm that they work in live or simulated environments
• Adjust code complexity, log volume, failure modes, and ambiguity to calibrate task difficulty
• Review and correct AI-drafted prompts and rubrics

Required Qualifications

You should be comfortable working with real software systems and explaining how engineering problems can be diagnosed, fixed, and evaluated. Strong written communication in English is essential because the role involves creating precise task instructions and correctness criteria.

• Hands-on software engineering experience in production or industry-level environments
• Strong debugging skills involving logs, observability tools, codebases, integrations, and configuration
• Experience tracing failures through code
• Fluency with Git, version control workflows, and code review practices
• Experience with APIs, services, infrastructure, configuration code, integrations, or CI/CD
• Ability to implement and verify code fixes
• Ability to define precise, checkable engineering outcomes
• Clear written and verbal communication in English

Helpful Background

The following experience is useful for creating varied, realistic engineering scenarios but is not stated as required.

• Automotive software
• Manufacturing
• Industrial IoT
• Enterprise SaaS
• Datadog, Grafana, or PagerDuty
• Embedded systems
• OT/IT integration
• SCADA environments

Why Work With OpenTrain

OpenTrain gives AI-training contributors a place to build a credible professional profile and grow a durable portfolio of work. By contributing to software engineering task evaluation, you can apply your technical judgment to how advanced AI systems are trained and assessed while working remotely with flexible hours.

• Build experience in the growing AI-training industry
• Showcase specialized engineering work in your OpenTrain profile
• Work remotely with a flexible part-time schedule
• Create tasks that influence how AI agents handle real engineering problems