Senior Software Engineer, Voice Agents

Health system call centers can't keep up with patient demand. 85% of patients who can't get through never call back. They go elsewhere, and appointments that should have been booked never are.

We're building AI Voice Agents that pick up every call and handle the full conversation: scheduling, rescheduling, cancellations, confirmations, FAQs, and intake. They work by integrating directly with the EHR and telephony stack in real time.

We're hiring a Software Engineer to build the systems these agents run on: the voice pipelines, EHR integrations, and services that take a call from "hello" to a booked appointment. It's a mid-level role on a small team that ships every day, with a lot of room to grow into harder problems.

This full-time position requires working 5 days a week in our Mountain View, CA office.

• Design and build backend services in Python that power real-time voice agents: call orchestration, conversation state, tool execution, and EHR write-back.
• Architect low-latency pipelines connecting telephony, ASR, LLMs, and TTS into a single conversational loop.
• Lead integrations with healthcare systems (like Epic, Cerner, athena, and others) for scheduling, eligibility, and intake.
• Drive design decisions on APIs, data models, and event-driven workflows, and mentor other engineers on the team.
• Partner with AI, product, and operations to ship features patient access teams actually use.
• Set the bar for what ships: latency, reliability, observability, and release readiness.
• 4+ years of backend software experience, primarily in Python.
• A track record of designing and shipping production services, APIs, and data models that others build on.
• An intuition for how people interact with automated assistants through voice and text, where the rules are different than icons and buttons.
• Deep experience with async, event-driven, or real-time systems.
• Strong fundamentals in testing, LLM evaluation, observability, and reliability.
• A patient-centered mindset. You design for the person on the other end of the line, not just the metrics.
• Bias for action, an ownership mindset, and the judgment to know when to move fast and when to slow down.