Staff AI Evaluation Engineer - Coding Assistant Quality

OpenKrill; United States; Remote; USD 138,000 - 276,000; Part time

Job summary

OpenKrill is seeking a Staff Software Engineer (AI Evals) in the United States. The contract role pays $100–$200/hour for ~20–40 hours per week, starting ASAP.

You will assess AI coding agents’ behavior, focusing on usefulness, reasoning quality, and engineering judgment rather than traditional production coding. Responsibilities include end-to-end evaluation of AI interactions, providing opinionated feedback, and helping define what great looks like when using Cursor and similar tools.

Qualifications

• Staff/Principal level engineering with strong judgment.
• Ability to evaluate AI-assisted coding interactions rigorously.
• Experience with modern AI-assisted dev workflows.
• Comfort giving direct, opinionated feedback to engineers.

Responsibilities

• Evaluate AI-generated coding interactions end-to-end.
• Judge usefulness, correctness, and engineering thinking.
• Provide clear feedback on what worked, didn’t, and what felt off.
• Distinguish quality levels in model outputs (2 vs 4).
• Help define what great looks like when using AI coding tools.

Skills

Go

Python

Rust

TypeScript / JavaScript

Tools

OpenAI Codex

Claude Code

Cursor

Job description

OpenKrill is seeking a Staff Software Engineer (AI Evals) in the United States. The contract role pays $100–$200/hour for ~20–40 hours per week, starting ASAP.

You will assess AI coding agents’ behavior, focusing on usefulness, reasoning quality, and engineering judgment rather than traditional production coding. Responsibilities include end-to-end evaluation of AI interactions, providing opinionated feedback, and helping define what great looks like when using Cursor and similar tools.