Founding engineer, voice agents

Sonorant; United States; Remote; USD 120,000 - 180,000; Full time

Benefits offered by this job

Equity

Remote-first

Health benefits

Paid time off

Job summary

Sonorant is hiring to build the core runtime and tooling for live voice agents in production. You will work on low-latency inference pipelines, policy-aware tool execution, and safety guardrails, shaping architecture and owning features end-to-end.

Ideal candidates communicate clearly, thrive in ambiguous problems, and have a track record in shipping production systems with observability and robust rollback capabilities. This role is remote-first with flexible hours across US/EU.

Qualifications

• Strong fundamentals in TypeScript/Node and async systems.
• Experience shipping and operating production systems (monitoring, debugging, rollbacks).
• Comfort owning ambiguous problems end-to-end with minimal direction.
• Clear written and verbal communication.

Responsibilities

• Design and implement real-time voice agent runtime components (VAD, STT/LLM/TTS orchestration, tool execution).
• Build and maintain evaluation harnesses for latency, quality, and safety.
• Define and enforce production safety patterns (timeouts, fallbacks, least-privilege tool scoping).
• Collaborate on API design, observability, and operational runbooks.
• Participate in code review, incident response, and on-call as the team scales.

Skills

TypeScript/Node

async systems

production monitoring

Job description

Sonorant helps teams build reliable, low-latency voice agents for real operations. We focus on containment, quality, and safety—with least-privilege integrations, auditable actions, and rollout plans tied to measurable outcomes.

About the role

You’ll build the core runtime and tooling that powers voice agents in production. This includes low-latency inference pipelines, policy-aware tool execution, evaluation harnesses, and production safety guardrails. You’ll work closely with the rest of the team to shape architecture and own features end-to-end.

Key responsibilities

• Design and implement real-time voice agent runtime components (VAD, STT/LLM/TTS orchestration, tool execution).
• Build and maintain evaluation harnesses for latency, quality, and safety.
• Define and enforce production safety patterns (timeouts, fallbacks, least-privilege tool scoping).
• Collaborate on API design, observability, and operational runbooks.
• Participate in code review, incident response, and on-call as the team scales.

Details

• Location: Remote (US/EU)
• Type: Full-time

Required qualifications

• Strong fundamentals in TypeScript/Node and async systems.
• Experience shipping and operating production systems (monitoring, debugging, rollbacks).
• Comfort owning ambiguous problems end-to-end with minimal direction.
• Clear written and verbal communication.

Preferred

• Experience with real-time audio, speech, or ML pipelines.
• Familiarity with CRM/ticketing/KB integrations or tool-calling agents.
• Prior work in a small or founding-stage team.

What we offer

• Competitive salary and equity.
• Remote-first; we support flexible hours and time zones (US/EU).
• Health, dental, and vision benefits; PTO and parental leave.
• Focus on impact and learning; we don't require a specific degree or years of experience.

Sonorant is an equal opportunity employer. We welcome applicants of all backgrounds.