Applied AI Engineer: Design Agent Evaluation & Optimization

present of AI; San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Job summary

Magic Patterns, an AI design platform that turns prompts and screenshots into production-ready UI components, seeks an engineer to own and enhance an evaluation system for a self-improving agent loop. You will focus on improving AI-generated design output and measure agent performance, while working with product to turn user feedback into improvements.

The role emphasizes building evaluation capabilities, post-training adjustments, and context engineering to guide model behavior in a

Qualifications

• Hands-on experience with AI models or agents.
• Experience building evals, post-training, or context engineering.
• Ability to think from first principles and translate user feedback into agent improvements.

Responsibilities

• Own an evaluation system that feeds a self-improving agent loop.
• Push the quality of AI-generated design output.
• Collaborate with product to translate user feedback into agent improvements.

Skills

AI model experience

Agent experience

Evaluation systems

First principles thinking

Job description

Magic Patterns, an AI design platform that turns prompts and screenshots into production-ready UI components, seeks an engineer to own and enhance an evaluation system for a self-improving agent loop. You will focus on improving AI-generated design output and measure agent performance, while working with product to turn user feedback into improvements.

The role emphasizes building evaluation capabilities, post-training adjustments, and context engineering to guide model behavior in a