ML Research Scientist, Audio Algorithms

Socket.dev; Mountain View (CA); On-site; USD 174,000 - 252,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

Google is seeking a high-velocity applied research engineer to lead exploratory ML-audio work focused on real-time perception for humans and AI. You will shape foundational algorithms and bridge theory with validated prototypes in an open-ended research setting.

The role emphasizes impact-first exploration, massive user differentiation, and collaboration with senior researchers to define next-generation agentic audio technologies. Equity and comprehensive benefits accompany the salary.

Qualifications

• PhD or equivalent in CS or related field and research leadership.
• Experience developing ML-audio algorithms for audio applications.
• Publications or submission track in top venues.

Responsibilities

• Develop new algorithms for real-time audio perception for humans and AI.
• Architect first-principles algorithms bridging theory and prototypes.
• Prioritize impact and user differentiation in research directions.
• Protect research velocity while guiding future-decade issues with execution teams.
• Collaborate with researchers to define foundational elements of agentic audio.

Skills

Audio ML research

Leadership experience

Real-time processing

Publications

Education

PhD in CS or related field

Equivalent practical experience

Tools

JAX

TensorFlow

C++

Python

Job description

Minimum qualifications:

• PhD in Computer Science, a related field, or equivalent practical experience.
• 2 years of experience leading a research agenda.
• Experience developing and training machine learning algorithms specifically for audio applications (e.g., neural noise suppression, acoustic modeling, or speech enhancement).
• Experience with deep learning frameworks such as JAX or TensorFlow applied to signal processing tests.
• One or more scientific publication submission(s) for conferences, journals, or public repositories (such as CVPR, ICCV, NeurIPS, ICML, ICLR, etc.).

Preferred qualifications:

• Experience in algorithm implementation and prototyping using C++ and Python.
• Experience taking theoretical ML-audio research ideas and implementing functional product proofs-of-concept.
• Experience navigating high-ambiguity research environments where the priority is delivering massive user differentiation over a specific method.
• Experience in developing foundational elements for ML based voice/audio algorithms.

About the job:

Join a new, high-velocity "startup-style" Applied Research team within TechEng dedicated to transformational audio bets for the next decade of Pixel and Buds. You will lead exploratory, non-timeline-based research into Open Ear Inteligibility and Superhuman Hearing for humans and machines.

We operate under a "Shielded but Connected" model—protected from the gravitational pull of daily product cycles to focus on pure feasibility and first-principles innovation. In this role, you will adopt an impact-first philosophy, ruthlessly prioritizing massive user differentiation to prove the "impossible" and define the foundational elements of agentic audio experiences. If you are an audio architect who grows in lean environments and wants to build the future of sound, this is your team.Individual pay is determined by factors including job-related skills, experience, and relevant education or training.

US: $174000 - $252000 (USD) + 15% bonus target + equity + benefits

Learn more about benefits at Google.

Responsibilities:

• Develop new and advanced algorithms for "moonshot" audio initiatives specifically real time perception for humans and AI operating outside the gravitational pull of immediate product timelines.
• Architect first-principles algorithms that bridge the gap between theoretical research and validated prototypes, proving feasibility before scaling.
• Adopt an impact-first philosophy, prioritizing massive user differentiation and maintaining the agility to pivot when technical paths do not yield high-order breakthroughs.
• Foster a "Shielded but Connected" environment, protecting the team's research velocity while providing technical guidance to core execution teams on future-decade issues.
• Collaborate alongside a small group of researchers (including Principal-level leadership) to define the foundational elements of next-generation agentic audio.