AI Conversation Designer

Zof AI; San Francisco, Northern (CA, KY); Hybrid; USD 110,000 - 160,000; Full time

Job summary

Zof AI is seeking an AI Conversation Designer to shape how our AI talks to people who depend on it. You’ll own tone, flow, and how it handles uncertainty, partnering with design, product, and AI engineering to craft a trustworthy interface.

Responsibilities include designing conversational behavior and escalation paths, testing system prompts with engineers, and iterating on transcripts to improve real-world conversations across surfaces.

Qualifications

• Experience in conversation design, UX writing, or content design.
• Working understanding of how LLM systems behave and fail.
• Excellent command of tone, clarity, and voice.
• Ability to prototype and test conversational flows.
• Comfort collaborating closely with engineers.
• Attention to detail and a high quality bar.
• Clear communication.
• Comfort operating in a fast-moving environment.

Responsibilities

• Design the conversational behavior of Zof AI products: tone, personality, and flow.
• Design how the AI handles uncertainty: when it hedges, asks, or escalates.
• Write and test system prompts and response guidelines with AI engineers.
• Design human-handoff moments so escalation feels safe, not broken.
• Review real transcripts and iterate on how conversations actually go.
• Partner with the evals team to make conversation quality measurable.
• Keep a consistent voice across product surfaces and channels.
• Raise the bar for how the product communicates under every condition.

Skills

Conversation design

UX writing

Content design

LLM behavior knowledge

Prototype testing

Engineer collaboration

Job description

Zof AI is seeking an AI Conversation Designer to shape how our AI talks to the people who depend on it: tone, flow, and above all how it handles uncertainty. When the AI is confident, unsure, or wrong, what it says next is a designed decision, and this role owns it. Sitting between design, product, and AI engineering, the ideal candidate treats language as an interface, tests their work against real transcripts, and knows that trust in an AI product is won or lost sentence by sentence.

Design · Mid-level · Full-time · On-site · San Francisco, CA

Responsibilities

• Design the conversational behavior of Zof AI products: tone, personality, and flow.
• Design how the AI handles uncertainty: when it hedges, asks, or escalates.
• Write and test system prompts and response guidelines with AI engineers.
• Design human-handoff moments so escalation feels safe, not broken.
• Review real transcripts and iterate on how conversations actually go.
• Partner with the evals team to make conversation quality measurable.
• Keep a consistent voice across product surfaces and channels.
• Raise the bar for how the product communicates under every condition.

Requirements

• Experience in conversation design, UX writing, or content design.
• Working understanding of how LLM systems behave and fail.
• Excellent command of tone, clarity, and voice.
• Ability to prototype and test conversational flows.
• Comfort collaborating closely with engineers.
• Attention to detail and a high quality bar.
• Clear communication.
• Comfort operating in a fast-moving environment.

Nice to have

• Experience designing for chatbots, agents, or voice interfaces.
• Experience with prompt engineering or system prompt design.
• Background in linguistics, writing, or human-computer interaction.
• Experience designing for trust, safety, or escalation flows.

Must understand how LLM systems behave and fail well enough to design around uncertainty