Deepgram

Staff Conversational Designer

San Francisco, CA, Remote. Posted a month ago

The role

You will define how Deepgram's voice agents converse, including persona, turn-taking, repair, confirmation, and pacing. You will work with ML and Engineering on behaviors like endpointing and barge-in while building conversational-quality evals and developer guidance.

What you will do

  • Define the persona and voice system for Deepgram voice agents
  • Design turn-taking, barge-in, and end-of-turn behavior with ML and Engineering
  • Design conversational repair, no-match and no-input handling, and confirmation strategy
  • Own latency-aware pacing and perceived responsiveness
  • Establish conversational-quality evals and a transcript review practice
  • Build reference agent experiences and developer-facing design guidance
  • Partner with ML and Research on ASR and TTS behavior
  • Set conversation-design principles, review standards, and shared vocabulary

Required qualifications

  • Deep experience designing conversational behavior for LLM-based voice agents or assistants
  • Real fluency with the speech pipeline, from ASR through LLM to TTS
  • Track record designing turn-taking, interruption, repair, and confirmation patterns that shipped in production
  • Evidence of building quality measurement into the practice, such as evals, transcript review, benchmarks, or failure-analysis loops
  • Exceptional writing craft, including sample dialogs, design guidance, and documentation
  • Experience influencing engineering and ML partners without authority over them
  • Experience designing for developers or technical users, including APIs, SDKs, and documentation surfaces
  • A working AI practice

Nice to have

  • Time on a named assistant or a production voice agent platform
  • Practice with Wizard of Oz testing and sample dialog methods
  • Hands-on work with eval tooling for LLM or voice quality
  • Experience in high-stakes or regulated conversation domains
  • Multilingual or cross-locale conversation design experience
  • Background in high-growth B2B companies with both self-serve and enterprise motions

About the company

Deepgram is a research company that productizes as fast as possible, providing speech models and real-time infrastructure to power AI systems. Their Voice Agent Platform includes speech-to-text, text-to-speech, and speech-to-speech via cloud APIs or self-hosted deployments.