Remote | Software Engineer (AI-Assisted Development) — $60–$80/hour

24-MAG LLC connects experienced professionals with remote consulting opportunities across technical, evaluation, and project-based workstreams. This part-time consulting role seeks experienced Software Engineers to evaluate AI-assisted software development sessions, review backend or full-stack code, assess debugging and development workflows, and provide rubric-based technical feedback.

Responsibilities:

• Evaluate complete software development sessions from initial requirements through implementation
• Assess whether resulting code is functionally and technically correct
• Review implementation choices for maintainability, reliability, and engineering quality
• Identify bugs, incomplete solutions, or problematic technical assumptions
• Apply practical software engineering judgement across realistic development scenarios
• Review backend or full-stack code for correctness and implementation quality
• Identify logical errors, regressions, edge cases, and debugging gaps
• Assess whether debugging approaches efficiently identify and resolve underlying problems
• Evaluate proposed fixes for completeness and technical soundness
• Distinguish substantive engineering issues from minor stylistic differences
• Evaluate coding sessions performed with AI-assisted developer tools
• Review how developers interact with coding assistants throughout implementation and debugging
• Assess agentic and specification-driven development workflows
• Identify ineffective prompting, unnecessary iteration, or poor use of available development context
• Evaluate whether AI-generated suggestions are appropriately verified before implementation
• Review multi-step coding trajectories rather than isolated code snippets
• Assess the sequence of investigation, implementation, testing, and refinement
• Determine whether development decisions follow a coherent and effective workflow
• Identify points where a stronger engineering approach should have been taken
• Evaluate both final outcomes and the process used to reach them
• Assess engineering decisions against established software development practices
• Evaluate architecture, implementation strategy, testing, and debugging choices
• Review whether assumptions are appropriately validated during development
• Identify unnecessary complexity or technically weak approaches
• Apply professional judgement to ambiguous or imperfect coding scenarios
• Assess development traces against structured project criteria
• Provide clear written explanations supporting evaluation decisions
• Identify specific evidence within the development session for each judgement
• Apply scoring and review standards consistently across assignments
• Distinguish correct but unconventional approaches from genuinely flawed implementations
• Work with development sessions involving tools such as Cursor, GitHub Copilot, Claude Code, or comparable AI-assisted coding environments
• Evaluate workflows involving repositories, specifications, code changes, debugging, and testing
• Assess how developers use tooling to investigate and modify existing systems
• Review interactions between automated assistance and human engineering judgement
• Contribute practical insight into effective developer-tool workflows

Requirements:

• 3+ years of professional software development experience
• Strong background in backend or full-stack software engineering
• Hands-on experience with AI-assisted coding tools such as Cursor, GitHub Copilot, Claude Code, or similar platforms
• Experience with agentic or specification-driven development workflows
• Strong code-reading and debugging skills
• Ability to evaluate multi-step software development trajectories for correctness and best practice
• Strong written communication and ability to provide precise technical feedback
• Fully remote within the United States
• Experience with Kiro or Amazon CodeCatalyst is preferred
• Previous experience evaluating or grading AI-generated code is advantageous
• Contributions to developer tooling are advantageous