← BACK TO CAREERS
Member of Technical Staff - AI Audio / Voice Engineer
San Francisco Bay Area
·
Full-time
·
180k-350k
·
level dependent
ABOUT NIRVA
Nirva is building the first AI fashion accessory for your mind — a necklace or bracelet that privately listens only to the wearer to enable proactive journaling, emotion tracking, and life coaching. Think Oura Ring for your Mind.
Backed by Lightspeed and South Park Commons, recognized at CES within 90 days of founding, and launching in the U.S. this fall. Co-founders bring backgrounds from Meta Reality Labs (Quest, Ray-Ban glasses) and fashion industry.
THE ROLE
We’re looking for an AI Audio / Voice Engineer to build the audio front door to Nirva’s intelligence layer: the systems that capture only the wearer, reliably, privately, and with as much processing on device as possible. This is the trust foundation for the whole product. If Nirva cannot tell who is speaking, reject everyone else, and do it in noisy real life without draining the battery or sending raw audio where it should not go, nothing built on top of it works. You will own voiceprint, on-device audio, and the technical systems behind our privacy promise, and work directly with the founding team to build them from the ground up.
WHAT YOU'LL OWN
Voiceprint: build speaker enrollment, voice embeddings, verification and identification so Nirva recognizes the wearer and separates them from bystanders, family, coworkers, and media
Tune and evaluate voiceprint in the real world: false accepts, false rejects, noisy cafes, walking, wind, cars, quiet rooms, illness, and voice drift over time, plus re-enrollment and recovery flows
Anti-spoofing and trust protections: replay and impersonation resistance, liveness signals where relevant, and clear fallbacks when confidence is low
On-device audio: design the on-device pipeline for voice activity detection, segmentation, speaker filtering, audio preprocessing, noise suppression, and streaming to speech recognition, with tight latency, power, thermal, and memory budgets
Decide the on-device vs. cloud split for each audio task and build graceful fallbacks when connectivity, battery, or confidence changes
Trust by design: on-device first processing, bystander exclusion, consent and recording controls, encryption in transit and at rest, data minimization, retention and deletion controls, and user-visible controls that match what the system actually does
Build evals and data pipelines for audio: test sets across people, accents, environments, and devices, failure analysis, and systematic debugging of real user-facing failures
Partner with product, design, and the memory / AI teams so clean, trusted, wearer-only audio becomes transcripts, memories, and experiences users can rely on
WHAT WE'RE LOOKING FOR
Must-haves
2+ years of full-time engineering experience, or a PhD in computer science, electrical engineering, machine learning, speech / audio, or a related field, plus 2 applied internships (industry or applied research) where you shipped or evaluated real systems
Deep, hands-on knowledge of at least one of these, and working knowledge of the other two: 1) voiceprint / speaker verification and identification, 2) on-device audio ML, 3) trust, privacy, and security for audio sensing
Hands-on experience with speech and audio ML in production: speaker embeddings, diarization, VAD, streaming audio, speech-to-text interfaces, audio preprocessing, or speech event detection
Experience running ML on device or at the edge: model quantization, distillation, latency and power optimization, and working within hardware constraints
Strong eval mindset for audio: you can build a test set, measure false accept / false reject and word error tradeoffs, find why a failure happened, and improve the system, not just the demo
Strong product engineering instincts and high ownership: you identify what needs to happen and drive it to shipped quality
Comfortable with ambiguous, human-centered problems where “good” is not easy to measure, and energized by 0-to-1 work, fast iteration, and early-stage startup intensity
Based in San Francisco and able to work in person
Strong signals (nice-to-haves)
Shipped always-on or low-power audio, wearables, hearables, or mobile audio products
Experience with anti-spoofing, voice liveness, or biometric security and its failure modes
iOS audio, Core ML, or embedded / DSP experience
Privacy engineering, on-device personal data systems, or consumer trust and safety for sensing products
Experience spanning audio ML, backend infrastructure, and user-facing surfaces
COMPENSATION
San Francisco Bay Area · Full-time · 180k-350k, level dependent
TO APPLY
Apply here
