Design and ship low-latency voice and conversational AI agent pipelines within Realm-X.

Architect and deliver voice and text agent pipelines for real-time, multi-turn customer interactions
Evaluate and implement principled trade-offs across frontier LLMs, smaller models, and routing strategies with a focus on reasoning depth vs latency
Lead a small pod of ML and platform engineers; improve agent evaluation, observability, and incident response
Work with Product and Voice channel teams to define KPIs, eval harnesses, and acceptance criteria for agent quality
Drive selective SLM fine-tuning and inference optimization to improve voice latency and cost

Deep, shipped experience with LangChain, LangGraph, LangSmith, and LangChain Deep Agents (or equivalent agent frameworks)
Hands-on with Voice-to-Voice models plus traditional TTS / STT pipelines; understands end-to-end voice models vs modular STT + LLM + TTS trade-offs
Strong grasp of LLM behavior for reasoning, tool use, structured output, and reasoning vs latency trade-offs across providers
Production experience with Twilio (or comparable telephony) and AWS
Expert Python, async programming, and WebSockets for real-time, bidirectional streaming
Solid foundation in deep learning, model evaluation, and inference optimization; deploy with Docker on AWS
Proven ability to lead a small team, mentor engineers, and partner credibly with Product and Design
Shipped production AI agents serving real users in voice and/or text channels
Think in pipelines and systems, not only models
Move fast, deliver impact, and maintain sound engineering judgment
Work with humility and collaboration; low-ego leadership and strong team elevation
Values work-life balance as a foundation for sustained high performance

Experience fine-tuning Small Language Models for domain-specific voice applications
Familiarity with RAG over structured business data and tool-using agents over API surfaces
Prior experience in regulated or customer-facing industries with strict reliability requirements
Publicly verifiable work on GitHub, in open-source agent frameworks, or in community competitions

Columbus, OH (onsite)
Late-stage candidates complete an in-person meeting with an AppFoliOian as part of the hiring process

Base pay range: USD 167,200 - 209,000 per year
Final compensation determined by skills, education, experience, and internal equity
Regular full-time employees are eligible for benefits
The listed compensation range excludes additional benefits and any discretionary bonuses
#LI-KB1

Sr. Machine Learning Engineer

Job Description