CX Conversational Designer for AI Support Agents

InvestedintheMission; United States; On-site; USD 85,000 - 100,000; Full time

Benefits offered by this job

Stock options

Remote-friendly policy

Job summary

Flock is seeking a dedicated conversation designer to own end-to-end design for our AI support agents. You will shape conversations from intent mapping to production deployment, ensuring the agent resolves issues accurately and knows when to hand off to humans.

Your work will focus on maintaining trust and reducing unnecessary escalations. You will collaborate with Support, Product, and Trust teams, building a analytics-driven backlog of fixes while upholding style, safety, and SLA alignment.

Qualifications

• Experience designing and iterating conversational flows for live AI agents.
• Ability to map intents, train utterances, and implement fallback strategies.
• Hands-on prompt development for LLM-driven conversations with performance data.

Responsibilities

• Own end-to-end conversation design for AI support agents from intent mapping to production deployment.
• Improve AI-resolved conversations while avoiding bad handoffs or escalations.
• Design escalation/fallback routing for SEV incidents, VIP, and Safe City accounts.
• Build a conversation analytics practice to turn data into backlogs of fixes.
• Create design standards and documentation to scale the agent portfolio without quality loss.
• Collaborate with Support, Product, and Trust to align agent behavior with SLA and safety.

Skills

Conversation design

Prompt development

LLM-based agents

Intercom/Fin platform

Performance dashboards

Tools

Intercom

Job description

Flock is seeking a dedicated conversation designer to own end-to-end design for our AI support agents. You will shape conversations from intent mapping to production deployment, ensuring the agent resolves issues accurately and knows when to hand off to humans.

Your work will focus on maintaining trust and reducing unnecessary escalations. You will collaborate with Support, Product, and Trust teams, building a analytics-driven backlog of fixes while upholding style, safety, and SLA alignment.