Backend Engineer, Cloud Infrastructure AI Evaluation

The work

You will build reinforcement learning environments for training and evaluating AI models. These environments must reflect realistic cloud operations and produce reproducible results.

You will combine backend engineering, cloud architecture, DevOps automation, and technical evaluation to test AI performance in production-style scenarios.

• Design environments that assess AI systems on infrastructure design, deployment, troubleshooting, security, scaling, and recovery.
• Create scenarios involving distributed systems, networking, identity and access management, message queues, durable storage, observability, rolling deployments, and disaster recovery.
• Develop deterministic validation tests and reference solutions for consistent assessment.
• Build defective variants and failure scenarios that test how models respond and recover.
• Document architecture, edge cases, and operational flows so environments are clear and reproducible.
• Refine environment specifications and acceptance criteria with technical collaborators.
• Use infrastructure automation and DevOps practices to deliver scalable, secure, maintainable evaluation systems.

What it pays and takes

This role is listed as entry level, but the work requires strong backend and cloud infrastructure skills. Prior AI experience is not required when you have the required backend and cloud expertise.

• Pay: $50 to $100 USD per hour.
• Time: 20+ hours per week.
• Work type: Part-time contractor.
• Language: English.
• Location: Applicants must be in a country included in the listing; this role is not marked worldwide.
• Programming: Strong experience with C++, Python, Rust, Go, Java, or JavaScript.
• Infrastructure: Practical experience with DevOps, cloud infrastructure, CI/CD pipelines, and automation tools.
• Systems: Ability to architect, scale, and secure distributed systems in production-grade environments.
• Technical knowledge: Networking, IAM, queues, durable storage, observability, deployments, and disaster recovery.
• Evaluation: Strong judgment when designing failure scenarios, acceptance criteria, and deterministic technical tests.

How it works

Apply on OpenTrain with your resume, then complete the application on the hiring site.

About AI training work

AI training work is the human work behind systems that learn from examples, including testing model behavior in realistic technical environments. People with strong specialist skills are needed to create reliable tests and judge whether an AI system handles complex tasks correctly.