Sunday (sunday.ai)

Software Engineer, Voice Interaction

Redwood City, CA

Sponsorship not specifiedDetected 61 days ago

C++Machine LearningLLMsAuditingRobotics

About the role

We're focused on an ambitious goal to make generalized robots broadly accessible, enabling households to take back quality time.
If you are ready to apply your skills to the forefront of robotics innovation, we'd love to hear from you.
Integrating machine learning models across local and cloud compute, you will transform raw audio signals into actionable instructions in a domestic environment.

Develop and maintain the full voice pipeline from microphone array input through wake word detection, speech-to-text, natural language understanding, and text-to-speech output
Design and optimize TTS output to deliver natural, responsive spoken interactions in real time on embedded hardware
Build reliable, well-tested software that runs on our robot, Memo, under real-world conditions including ambient noise, partial utterances, and unreliable network connectivity
Deliver a successful voice interaction experience to our Beta users
2+ years experience developing voice-driven systems including speech-to-text, text-to-speech, and real-time audio processing, with at least one end-to-end pipeline shipped to users or deployed on hardware

Strong understanding of both classical decision-making approaches (state machines, behavior trees, planning) and modern ML-driven reasoning (LLMs, VLMs)
Proficiency in C++, with experience in asynchronous programming, streaming/buffering patterns, and integration with cloud API services

Experience as a founding or early hire
able to define a release roadmap where no blueprint exists
Experience shipping responsive AI systems in robotics, video games, or embodied AI
Practical experience (or high curiosity) in interfacing with multimodal models
Publications in multimodal models, audio interpretation, or robotics
Even if you don't meet every single requirement, we encourage you to apply.
Studies show that women and underrepresented groups often hold back unless they meet 100% of the criteria - we don't want that to be the reason we miss out on great talent.

This listing is sourced directly from Sunday (sunday.ai)'s careers page and normalized into a canonical job model.