Interaction Designer — Engineer-Focused AI Review UX

Zof AI; San Francisco, Northern (CA, KY); Hybrid; USD 90,000 - 120,000; Full time

Job summary

Zof AI is seeking an Interaction Designer to design how engineers come to trust autonomous agents that change their code. This role owns the research and interaction work behind the review surface: reading the evidence behind a proposed fix, judging whether the validation holds, and approving it or sending it back.

The ideal candidate talks to real engineers, turns those conversations into flows and interaction patterns, and treats the design system as how good decisions stay consistent.

Qualifications

• Experience designing complex product flows end to end.
• Portfolio showing research that changed a design decision.
• Working knowledge of research methods, from interviews to usability testing.
• Ability to move from research to wireframes to interactive prototypes.
• Familiarity with design systems and component-level thinking.

Responsibilities

• Design the review experience where engineers accept, reject, or revise agent-proposed code changes.
• Run research with engineers and QA leads to learn how trust in automation is built and lost.
• Design how evidence, reproduction steps, and validation results are presented so review decisions are fast and informed.
• Turn research findings into flows, wireframes, and interaction specs engineering can build from.
• Prototype and test interaction patterns for agent status, uncertainty, and failure states.
• Extend the design system with components and patterns for reviewing machine-generated work.
• Partner with product and engineering from problem framing through shipped experience.

Skills

End-to-end design

UX research

Wireframes to prototypes

Design systems

Engineer collaboration

Clear communication

Job description

Zof AI is seeking an Interaction Designer to design how engineers come to trust autonomous agents that change their code. This role owns the research and interaction work behind the review surface: reading the evidence behind a proposed fix, judging whether the validation holds, and approving it or sending it back.

The ideal candidate talks to real engineers, turns those conversations into flows and interaction patterns, and treats the design system as how good decisions stay consistent.