Project Summary

This project focuses on evaluating and improving AI-generated React applications. Consultants will create realistic natural language specifications, design structured evaluation rubrics, assess AI-generated apps for functional accuracy and UX/UI quality, and run standardized evaluations across multiple AI models. Work involves identifying model failure modes, documenting comparative results, and producing recommendations that improve AI agents’ ability to generate high-quality applications from natural language inputs.

This is a project-based consultant role. Consultants will be paid on a per-project basis; hourly rates are estimates based upon anticipated time of completion. Consultants will control their own schedule, provide their own tools to perform the work, and are free to simultaneously provide services to other vendors and employers to the extent permitted by those vendors and employers.

Responsibilities

Contributors will:

• Create realistic natural language prompts describing functional app requirements
• Design weighted evaluation rubrics covering functionality, UX/UI quality, responsiveness, accessibility, and error handling
• Assess AI-generated React applications against specifications and rubrics
• Run standardized evaluations across multiple AI models and document comparative results
• Identify failure modes, strengths, and improvement opportunities in AI outputs

Expected Outcomes

• Completed prompts, rubrics, and structured evaluation reports
• Comparative assessments across multiple AI models with identified improvement opportunities

Qualifications

• Experience working as a No-Code Developer, Product Designer (UX), Frontend Developer (React), Retool/Airtable Apps Builder or Solutions Analyst
• 3+ years building internal tools, websites, or applications
• Experience with React or no-code platforms (Airtable, Glide, Retool, Bubble)
• Ability to translate natural language specs into app flows and judge functionality against requirements
• Strong UI/UX sensibility and understanding of accessibility standards (ARIA, responsive design)