Frontend Code Evaluation AI Trainer

Software Development & AI Training • Independent Contractor • Remote • Weekly Payouts

JobHub by NeonLabs, in partnership with Mercor, is seeking experienced frontend and full-stack software developers across eligible global regions to support leading AI labs in training frontier models on frontend code evaluation.

This role focuses on leveraging your web development expertise to evaluate and grade AI-generated frontend code against real-world reference web pages. You will apply your deep understanding of hand-written HTML, CSS (flexbox, grid, legacy floats/tables), browser DevTools, and responsive design to judge visual fidelity, structure, and code construction quality.

What You'll Do

• Render reference web pages and candidate replications at a 1920×1080 viewport to judge closer reproductions state-by-state
• Diff visual fidelity in detail: box model and spacing, typography, color and borders, asset handling, z-order, and overflow
• Read source code of model attempts to grade construction quality — distinguishing genuine correctness from superficial rendering (spotting hardcoded offsets, absolute positioning standing in for layout, inline style soup, etc.)
• Test responsiveness and write specific, evidence-cited justifications for every preference
• Collaborate remotely with leading AI researchers to shape code-generation AI systems without utilizing confidential or proprietary employer information

Ideal Qualifications

• 3–8 years of professional web development experience shipping web interfaces for a living (frontend or full-stack)
• Fluency across web eras: command of hand-written HTML/CSS, semantic markup, modern layout (flexbox, grid), and legacy float- and table-based layouts
• Browser DevTools as muscle memory and command-line comfort for standing up local static servers and unzipping site trees
• At least one completed Mercor engagement, delivered in full — net-new experts are not onboarded to this project
• Access to a desktop or laptop with a 1920×1080 viewport and administrator rights to run a local server
• Professional written English for drafting clear, evidence-based task justifications

What Makes This Role Unique

• Work directly alongside leading AI labs on frontier code-evaluation and model training
• Fully remote role that can be completed on your own schedule (units take roughly 2–3 hours and are timed)
• Opportunity to shape the next generation of AI coding assistants and code evaluation benchmarks
• Competitive compensation with weekly payouts via Stripe or Wise

Details

• Pay: Based on services rendered, with weekly payouts
• Type: Independent Contractor • Remote • Platform: Mercor (weekly via Stripe or Wise)
• Location Requirement: Open across eligible regions (including United States, Canada, United Kingdom, EU member states, and Latin America)
• Schedule: Flexible schedule requiring contiguous multi-hour blocks (~2–3 hours per unit)
• Engagement Details: Task-based code evaluation work requiring established Mercor history; projects can be extended, shortened, or concluded early depending on needs and performance

Submit your application via the link below. Qualified candidates with prior completed Mercor engagements may move through the selection process and conduct a quick AI interview.