Job description

ABOUT PHONIC Phonic is an audio research lab focused on powering the most realistic, human-like voice AI conversations. We’ve rethought the entire stack in pursuit of this goal, from models to product, to create voice agents that feel like they truly understand you, respond emotionally, and perform agentic tasks with frontier intelligence. Our customers build voice-native AI products in industries like customer support, healthcare, logistics, and recruiting. We’ve grown ~20x since the start of this year, and we’ve raised over $30M from tier-1 VCs, including First Round Capital, Lux Capital, and Felicis. About the Team Phonic has an extremely talent-dense and close-knit team of top-tier AI researchers, international olympiad medalists, and former founders. We work with high trust and keep raising the bar on how we deliver world-class research and product experiences. Everyone takes ownership of what they do and isn’t afraid to dive headfirst into new problems. We’re fully in person in our SF office. ABOUT THE ROLE As a Research Engineer at Phonic, you'll sit at the intersection of cutting-edge ML research and production engineering, working directly on the core systems that make Phonic's voice AI feel genuinely human. You'll design, train, and iterate on models across the voice stack (speech, audio, language, and beyond), while also building the infrastructure and tooling needed to move fast from research idea to deployed product. This is a role for someone who thrives in ambiguity, moves with urgency, and takes full ownership of problems. You'll work closely with researchers and product engineers in a high trust, in-person environment in our SF office to push the frontier of what voice AI can do. WHAT YOU’LL DO - Design, implement, and iterate on models across the voice stack from audio and speech to language and beyond - Build the training pipelines, evaluation frameworks, and tooling that let us experiment and iterate quickly - Translate research results into production-grade systems alongside our engineering team - Form hypotheses, chase down interesting results, and take ownership over the problems you work on WHAT YOU’LL BRING - Hands-on experience in ML research or research engineering, industry or PhD-level academia - Proficiency in Python and PyTorch (or JAX), and the ability to implement models cleanly from papers - Experience running ML experiments end-to-end: data processing, training, evaluation, and iteration - Comfort in fast-moving, ambiguous environments - defining the problem is part of the job - The ability to clearly explain what you tried, what worked, and why NICE TO HAVE - Research experience in speech, audio, or language modeling (ASR, TTS, LLMs, codec models) - Familiarity with diffusion, flow matching, or autoregressive generative models - Experience with distributed training, quantization, or inference optimization - Competitive programming or olympiad background BENEFITS - 💸 Top-tier compensation: in order to get the best talent, we provide salary and equity that recognize your skillset - 🥗 Meals: free breakfast, lunch, and dinner provided in the office - 🩺 Healthcare: Comprehensive health, dental, and vision - 🤝 We have regular off-sites and team celebrations - 👵 401(k) – Let us help you plan for the future. We’ve got you covered.

Requirements

Must-haves

Design, implement, and iterate on models across the voice stack from audio and speech to language and beyond
Build the training pipelines, evaluation frameworks, and tooling that let us experiment and iterate quickly
Translate research results into production-grade systems alongside our engineering team
Form hypotheses, chase down interesting results, and take ownership over the problems you work on
Hands-on experience in ML research or research engineering, industry or PhD-level academia
Proficiency in Python and PyTorch (or JAX), and the ability to implement models cleanly from papers
Experience running ML experiments end-to-end: data processing, training, evaluation, and iteration
Comfort in fast-moving, ambiguous environments - defining the problem is part of the job
The ability to clearly explain what you tried, what worked, and why

Nice-to-haves

Research experience in speech, audio, or language modeling (ASR, TTS, LLMs, codec models)
Familiarity with diffusion, flow matching, or autoregressive generative models
Experience with distributed training, quantization, or inference optimization
Competitive programming or olympiad background

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aimlpython

Platform: ashby
Effort: quick

Benefits

Benefits detected in the listing.

Dental

Vision

401(K)