Problem Design Engineer

We create high-quality coding problems for training AI models — drawn from real production repositories, not synthetic puzzles. Your role is to ensure these problems are fair, clear, and technically sound.

Part-Time Code Quality Reviewer | Remote | $40/hr | 0-25 hrs/week

We’re looking for engineers who can trace the intent behind a code change, understand what changes were made and why, and express that clearly in natural language.

This is judgment work, not mechanical testing. You'll make philosophical decisions about fairness and clarity in technical contexts.

• Review code changes across unfamiliar repositories and tech stacks

• Assess problem fairness: Could a skilled engineer solve this with only the given context?

• Evaluate test quality: Do tests properly validate the intended solution?

• Write clear justifications for your assessments (2-3 sentences per decision)

• Flag coherence issues: Inconsistent artifacts, unclear requirements, missing context

Time commitment: 1-4 hours per problem, fully async — work whenever fits your schedule.

Ideal candidates:

• PhD students in CS/computational fields seeking long-term flexible income

• Open source contributors who care deeply about code quality

• Undergrad with at least 1 year in CS, ECE, Computational Biology, or related technical programs

• Professional engineers seeking intellectually interesting part-time work

You'll thrive if you:

• Have strong software engineering fundamentals

• Can work independently with minimal supervision

• Write clearly and precisely

• Enjoy making judgment calls in ambiguous situations

• Stay focused during 1-3 hour review sessions

• 1+ years of software engineering experience (internships, research projects, or professional work)

• Fluent English; strong writing is a must

• Responsive, independent, and attentive to detail

• 0–25 hours per week, flexible

• $40/hour for all review work

• Performance-based advancement: Top contributors can receive promotions and bonuses

• Long-term engagement: We're building a scalable team, not a temporary workforce

1. 15-minute async video interview — answer 4 questions about your technical background and work style

2. Sample problem review — demonstrate your assessment skills on a real coding problem

3. Onboarding — join our active problem design engineering community and start reviewing

Q: Is this suitable for PhD students?
Yes — many of our top contributors are PhD students. The work is fully async and fits around research schedules.

Q: Do I need to be available during specific hours?
No. Work whenever you want. Some problem design engineers work weekends, others weekday evenings.

Q: What if I'm not familiar with the programming language in a problem?
That's expected. You're assessing problem fairness and clarity, not implementing solutions. Familiarity with general software engineering principles is what matters.

Q: How quickly can I start?
If you pass the video interview and sample review, you can start within 1-2 weeks.

Q: Is this a path to full-time employment?
This is a long-term contract role, not a pipeline to full-time positions. We value contributors who excel at this specific type of work.

The Widget Factory is the quality assurance division of Idler (YC S25), where we ensure coding problems meet research-grade standards for AI training. Our mission is to build the most reliable QA system for AI coding benchmarks.