Voice Agents AI Engineer - Real-Time Speech & Calls

Rovrstack; New York (NY); On-site; USD 170,000 - 230,000; Full time

Job summary

RoVR in New York, NY seeks an Applied AI Engineer (Voice Agents) to build autonomous voice agents that talk to patients, verify identity, schedule, and hand off to staff. You will own end-to-end real-time speech and evaluation loops, working across Senior and Staff levels to reduce latency and improve reliability.

You will collaborate with clinics and PMs to shape what constitutes a great call, optimize prompts and tools, and ensure PHI is handled securely.

Qualifications

• 5+ years of software engineering experience with hands-on shipping of LLM-powered products.
• Strong Python, including async, and experience with LLM APIs.
• Experience with real-time voice systems: WebRTC, LiveKit or similar, STT, and TTS.
• Comfortable using evals to drive quality.
• Iterative mindset: build proofs of concepts, decide quickly, and ship v0s.
• Nice to have telephony (SIP, Twilio) or experience with HIPAA and PHI.
• RoVR is an equal opportunity employer and does not discriminate on basis of race, gender, etc.

Responsibilities

• Build and ship voice agents end to end: real-time speech pipelines, turn-taking, interruptions, and call outcomes.
• Partner with clinics and PMs to define what a great call looks like, and turn it into evals.
• Optimize performance through prompt engineering, model selection, tool design, and context management.
• Drive latency and cost down with smart model routing, caching, and streaming.
• Build observability for calls (transcripts, traces, metrics) with PHI redacted.
• Stay current on speech and agent models, and bring what works into production.

Skills

Software engineering

Python (async)

LLM APIs

Agent frameworks

Real-time voice

WebRTC

LiveKit

STT

TTS

Evals

Prompt engineering

Latency optimization

Tools

Twilio

SIP

Job description

RoVR in New York, NY seeks an Applied AI Engineer (Voice Agents) to build autonomous voice agents that talk to patients, verify identity, schedule, and hand off to staff. You will own end-to-end real-time speech and evaluation loops, working across Senior and Staff levels to reduce latency and improve reliability.

You will collaborate with clinics and PMs to shape what constitutes a great call, optimize prompts and tools, and ensure PHI is handled securely.