Software Engineering AI Task Author

OpenTrain AI; Canada; Hybrid; CAD 135,000 - 231,000; Part time

Benefits offered by this job

Part-time engagement

Pay range: $70-$120 USD per hour

Location: Canada

Job summary

OpenTrain AI is hiring a Software Engineering AI Task Author to design and calibrate engineering challenges used to train and evaluate AI agents. You will create scenarios with integration failures, log-based debugging, and code-tracing tasks for real-world validation.

The role blends hands-on software engineering with evaluation design, requiring clear criteria and measurable outcomes to ensure tasks are authentic and appropriately challenging.

Qualifications

• Hands-on software engineering experience in industry or production environments.
• Ability to move from symptoms to root cause and implement fixes.
• Experience with logs, observability, and debugging.

Responsibilities

• Author task prompts based on realistic production-style failures and debugging workflows.
• Build task environments with codebases, logs, alerts, and system states that represent plausible engineering situations.
• Write scoring rubrics defining correct diagnoses and fixes, including validation criteria where applicable.
• Solve each task yourself, implement the fix, and verify that the expected outcome works.
• Adjust code complexity, log volume, failure modes, and ambiguity to calibrate task difficulty.
• Review and correct AI-drafted prompts or rubrics.

Skills

Software engineering

Debugging

Git/version control

Code review

CI/CD

Observability

Infrastructure experience

Job description

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help contributors discover specialized projects, build a professional profile, and apply their skills to work shaping the next generation of artificial intelligence.

As an OpenTrain contractor, you can turn practical engineering expertise into a focused AI training portfolio. Creating an account is free, and your profile can help demonstrate credible experience across this growing field.

About AI Training Work

AI systems learn and improve from examples prepared, reviewed, and evaluated by people. In software engineering projects, experienced contributors create coding tasks, assess model-generated solutions, and define the criteria that distinguish a working fix from an incomplete or incorrect one.

This is cutting-edge technical work at the human side of AI development. Many AI training projects offer flexible, part-time schedules, making it possible to contribute alongside other professional or personal commitments.

The Role

OpenTrain is seeking a Software Engineering AI Task Author to design and calibrate realistic engineering challenges used to train and evaluate advanced AI agents. You will create scenarios involving integration failures, log-based debugging, source-code bug tracing, configuration changes, and verified fixes.

The work combines hands‑on software engineering with careful evaluation design. Each task should feel authentic, have an appropriate level of difficulty, and include outcomes that can be checked objectively.

• Contractor, part-time engagement
• Time requirement: 20+ hours per week
• Pay range: $70-$120 USD per hour
• Location: Canada
• Primary language: English
• Listed experience level: Entry level

What You’ll Do

You will build complete engineering scenarios and validate them yourself before they are used for AI training or evaluation. Your work will cover task authoring, environment setup, scoring design, debugging, and quality review.

• Author task prompts based on realistic production-style failures and debugging workflows.
• Build task environments with codebases, logs, alerts, and system states that represent plausible engineering situations.
• Write scoring rubrics defining correct diagnoses and fixes, including validation criteria where applicable.
• Solve each task yourself, implement the fix, and verify that the expected outcome works.
• Adjust code complexity, log volume, failure modes, and ambiguity to calibrate task difficulty.
• Review and correct AI‑drafted prompts or rubrics.

Required Skills

This role requires practical experience working as a software engineer in industry or production environments. You should be comfortable moving from symptoms to root cause, implementing a fix, and establishing precise evidence that the result works.

• Hands‑on software engineering experience with infrastructure, integrations, or configuration code.
• Strong debugging ability, including reading logs, using observability tools, and tracing failures through source code.
• Fluency with Git, version control workflows, and code review practices.
• Experience with infrastructure, integrations, APIs, services, configuration files, or CI/CD systems.
• Ability to write and verify fixes in live or simulated environments.
• Ability to define precise, checkable criteria for engineering outcomes.

Helpful Background

The following experience is useful but is not identified as required. It can help you create more realistic scenarios across different technical environments and operational contexts.

• Experience in enterprise SaaS, e‑commerce, telecommunications, or industrial engineering.
• Familiarity with monitoring and observability platforms such as Datadog, PagerDuty, or Grafana.
• Background with cloud‑native architectures on AWS, GCP, or Azure.

Why Work in AI Training

Every major AI system depends on human contributors who prepare examples, evaluate outputs, and define what high‑quality performance looks like. Your engineering judgment can directly influence how advanced AI agents diagnose problems, modify code, and verify solutions.

OpenTrain provides a way to build a durable AI training portfolio around specialized work. By documenting projects and developing a stronger profile, you can show credible experience and discover future opportunities aligned with your technical skills.