
Deepgram
Staff Conversational Designer at Deepgram
$180,000–$240,000 / year
Apply for this positionRole overview
An at-a-glance view of the facts included with this Staff Conversational Designer at Deepgram listing.
- Company
- Deepgram
- Location
- Remote, United States
- Work arrangement
- Remote
- Employment type
- Full-time
- Compensation
- $180,000–$240,000 / year
Applications for this listing are submitted through CareersForge. Use the form on this page, then sign in to your Candidate Portal to follow application updates and any interview invitation.
Role overview
Deepgram is hiring a Staff Conversational Designer to define how its voice agents communicate, including persona, turn-taking, repair, confirmations, and pacing. This Staff individual-contributor role reports to the Director of Product Design and focuses on creating patterns and guidance that support both Deepgram’s experiences and customers building on the Voice Agent API.
You’ll work with ML, Research, and Engineering to translate conversation goals into production behavior, balancing responsiveness with interruption risk and real-time latency constraints. The role also establishes conversational-quality evaluations and transcript review practices, then uses production failures to improve the design system. Initial priorities are a documented persona and voice system and turn-taking and repair behavior designed and shipped with Engineering; the scope then grows to include quality evaluations and developer-facing guidance.
The position is remote within the United States, based in Pacific Time, with a preference for candidates in San Francisco. Deepgram states that expertise and leadership matter more than years of experience. Compensation is listed at $180,000–$240,000 per year.
Role at a glance: Staff Conversational Designer defining voice-agent behavior, quality evaluations, and developer guidance for Deepgram’s Voice Agent API.
Responsibilities mentioned in the listing
- Define and maintain a coherent persona and voice system across agent experiences and use cases.
- Design turn-taking, barge-in, and end-of-turn behavior with ML and Engineering, balancing responsiveness against the risk of interrupting users.
- Create conversational repair, no-match, no-input, and confirmation patterns, including safeguards for high-stakes actions.
- Set latency-aware pacing practices for brevity, backchanneling, hold speech, and filler speech.
- Establish conversational-quality evaluations and transcript reviews that turn production failures into a repeatable design process.
- Create reference agent experiences and developer-facing guidance for building conversations with the Voice Agent API.
- Partner with ML and Research on ASR and TTS behavior and criteria for conversation quality.
Qualifications stated in the listing
- Deep experience designing behavior for LLM-based voice agents or assistants, beyond scripted IVR flows.
- Fluency across the speech pipeline, from ASR through LLM to TTS, and understanding of where design decisions are implemented.
- A record of shipping turn-taking, interruption, repair, and confirmation patterns that perform in production.
- Evidence of incorporating quality measurement into design practice, such as evaluations, transcript review, benchmarks, or structured failure analysis.
- Excellent writing, including sample dialogs, design guidance, and documentation others can use.
- Experience influencing Engineering and ML partners on behavior they own, without direct authority.
- Experience designing for developers or technical users, including APIs, SDKs, and documentation.
- A working AI practice and a considered perspective on where AI tools help and where they can mislead.
