Coding Puzzle Solver for AI Evaluation

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. It helps contributors discover opportunities, build a credible AI training profile, and grow flexible project work into a long-term portfolio.

Creating an OpenTrain account is free, and candidates can apply through OpenTrain for remote work that matches their technical skills.

• Build a unified portfolio of AI training and data-labeling experience
• Discover projects aligned with your programming and evaluation skills
• Apply for flexible remote contractor opportunities

About AI Training and Code Evaluation

AI training is the human side of building artificial intelligence. People prepare examples, test model behavior, and review outputs so modern AI systems can produce more accurate, reliable, and useful results.

In this role, you will work on the software-focused side of AI evaluation by solving programming challenges and checking whether AI-generated solutions are correct, complete, reproducible, and compliant with technical requirements.

• Contribute directly to the evaluation of cutting-edge AI models
• Use programming expertise to identify subtle logic and edge-case failures
• Work remotely with flexible scheduling

The Coding Puzzle Solver Role

OpenTrain is recruiting a Coding Puzzle Solver to create and evaluate reproducible software solutions for AI training. You will tackle self-contained algorithmic, debugging, logic, and technical reasoning challenges that test advanced AI models.

The work may involve unfamiliar codebases and terminal environments. You will check solutions against explicit or hidden constraints and review AI-generated code and reasoning for flawed logic, incomplete explanations, edge-case failures, and technical requirement violations.

• Role type: Part-time contractor
• Location: Worldwide, fully remote
• Schedule: Approximately 15 hours per week
• Experience level: Entry level
• Language: English
• Compensation listed: $50-$80 per hour

What You'll Do

You will turn precise or ambiguous technical requirements into implementable solutions, then create clear evidence that those solutions work. Your results should be reproducible and suitable for automatic execution and evaluation.

The role combines independent problem solving, software development, debugging, testing, and careful review of AI-generated solutions.

• Solve coding, logic, debugging, and technical reasoning challenges
• Select appropriate algorithms, data structures, and problem-solving strategies
• Implement efficient solutions in a suitable general-purpose programming language
• Work in terminal-based development environments and existing codebases
• Debug code, tests, dependencies, and environment configuration issues
• Develop tests and verification methods that distinguish correct solutions from superficially plausible ones
• Review AI-generated code and reasoning for errors, missing cases, and constraint violations
• Produce clear, reproducible results that can be executed and evaluated automatically
• Explain implementation decisions, reasoning, and trade-offs clearly

Required Skills

This position requires strong algorithmic problem-solving ability and the confidence to investigate unfamiliar technical problems independently. You should be able to write testable software, reason about correctness and complexity, and communicate technical decisions clearly.

• Proficiency in at least one of C++, Python, Java, Rust, Go, C#, JavaScript, or TypeScript
• Strong understanding of common data structures, algorithms, complexity, edge cases, and correctness
• Experience debugging unfamiliar code or solving technically ambiguous problems
• Ability to work independently in terminal-based environments
• Ability to design verification methods for software solutions
• Ability to identify logic errors, incomplete reasoning, and constraint violations in AI-generated solutions
• Ability to produce reproducible software solutions
• Clear written and verbal technical communication

Helpful Background

The following experience is helpful but is not listed as required. It can demonstrate practical algorithmic problem-solving and experience evaluating or producing high-quality code.

• A public or otherwise verifiable Codeforces profile showing meaningful competitive-programming experience
• Experience participating in or judging competitive C++ events
• Open-source programming experience
• Recognized competitive-programming achievements

Schedule and Compensation

This is a flexible, global remote contractor role requiring approximately 15 hours per week. You may choose the hours and days you work, including weekends if desired.

Compensation is listed at $50-$80 per hour. The role also describes output-based task compensation, so the applicable compensation structure should be confirmed before work begins.

• Approximately 15 hours per week
• Choose your own working hours and days
• Weekend work is permitted if desired
• Global remote availability
• Contractor engagement
• Confirm the applicable compensation structure before starting work