Software Engineer, AI Research & Evaluation (US)

OpenTrain AI, Inc.; United States; Remoto; USD 69.000 - 207.000; Part-time

Descrizione del lavoro

OpenTrain AI, Inc. is seeking a Software Engineer for AI Research & Evaluation. The role focuses on building Python tools to verify code and assess AI-generated solutions across backend, data, ML infrastructure, and full-stack projects, as a remote US contractor.

You will review AI-generated code for quality, reliability, and performance, and create checks for engineering tasks while explaining technical decisions clearly. Flexible 20+ hours weekly, remote within the United States.

Competenze

• Strong Python skills, including production frameworks and engineering practices.
• Experience building full-stack applications and deploying scalable software.
• Knowledge of software architecture, design, development, debugging, and code review.

Mansioni

• Review code for backend services, data pipelines, ML infrastructure, and full-stack apps.
• Build Python agents and automated checks to verify solution quality.
• Evaluate AI-generated code for quality, reliability, and performance across projects.
• Explain technical decisions clearly in writing and conversation.
• Create checks for software engineering tasks and compare with industry benchmarks.

Conoscenze

Python

Code review

Full-stack

Backend

Testing

Formazione

CS fundamentals

Strumenti

Python tooling

Descrizione del lavoro

Software Engineer, AI Research & Evaluation

Build Python tools to verify code and assess AI-generated solutions across backend, data, machine learning infrastructure, and full-stack projects. This remote US contractor role calls for 20+ hours per week.

The work

Create and refine coding examples and solutions used to train and evaluate AI models. Review AI-generated code for quality, reliability, scalability, and performance across software projects.

• Review code for backend services, data pipelines, machine learning infrastructure, and full-stack applications.
• Build Python agents and automated checks to find errors and verify solution quality.
• Assess work across prototyping, architecture and API design, implementation, launch, testing, monitoring, and maintenance.
• Create checks for software engineering tasks and compare solutions with industry performance benchmarks.
• Explain technical decisions clearly in writing and conversation.

What to expect and bring

• Location: Remote, open to candidates based in the United States.
• Schedule: Flexible commitment of 20+ hours per week.
• Engagement: Part-time contractor role, initially one month, with possible extensions based on performance and fit.
• Strong Python skills, including production frameworks, tools, and engineering practices.
• Experience building full-stack applications and deploying scalable software.
• Knowledge of software architecture, design, development, debugging, and code review.
• A computer science foundation and experience with production systems, backend services, data pipelines, machine learning infrastructure, automated testing, or enterprise software are helpful.

About AI training work

AI training work uses examples prepared and reviewed by people to help AI systems learn; code evaluation is one way contributors assess and improve model outputs. OpenTrain is a free platform where people can find AI training projects, build a profile, and apply.

Requirements

Evaluate coding agents on real software tasks, run and review their code, and explain defects. This US contract offers up to $150 per hour, remote work, and flexible scheduling.

Create realistic game development tasks, review AI-generated work, and explain technical issues clearly. This remote, part-time project offers up to $125 per hour with no set schedule or minimum hours.

Design and review frontend coding tasks based on screenshots, mockups, diagrams, and recordings, and build reference implementations. The work pays approximately $640 to $800 per approved task.

Build research-grade programming tasks and assess AI-generated code in areas such as compilers, type systems, and verification. This contractor role pays $50 per hour for 20+ hours a week and requires a completed or in-progress PhD.

Review real engineering changes for bugs, security risks, design, and maintainability at $150 per hour. Remote freelance work takes about 30 minutes per review, with 5 to 20 hours of work per week.