Senior Machine Learning Research Engineer

About the role

The team bridges cutting-edge research with scalable services. You will own the design and execution of audio-centric machine learning initiatives that directly shape product strategy and user value. This role requires you to translate ambiguous problems into structured research questions and high-confidence solutions. You will define the technical roadmap for audio intelligence, balancing innovation with reliability and deployment constraints. You will act as a thought partner to both engineering and product teams, ensuring that audio ML capabilities align with business objectives. A core part of your work will involve turning novel signal processing ideas into robust services that handle real-world data complexity. You will establish best practices for model evaluation and data quality specific to the audio domain. Your contributions will be visible in the performance and stability of production systems serving large-scale audio workloads.

Key facts

What you'll do

Design and implement advanced audio signal processing pipelines that transform raw waveform data into high-fidelity training and inference signals.

Develop and optimize machine learning models, focusing on novel architectures and algorithms tailored to the temporal and spectral nature of audio data.

Collaborate closely with Operations to construct and maintain large-scale evaluation datasets that drive measurable improvements in model accuracy and robustness.

Architect end-to-end systems that ensure durable, low-latency inference across terabyte-scale audio corpora in production environments.

Conduct experiments to diagnose model behavior, isolate failure modes, and iteratively refine audio representations and learning objectives.

Lead the definition and execution of ML roadmaps, prioritizing research milestones and infrastructure investments that de-risk long-term audio capabilities.

Own model quality assessment frameworks that balance user experience metrics with business impact, ensuring alignment between technical outcomes and product goals.

Convert insights from reading research papers into production-ready code, maintaining a sharp focus on scalability, maintainability, and performance in cloud deployments.

Deploy and monitor ML systems for cloud-based inference, implementing observability and reliability practices specific to audio workloads.

Champion best practices for data curation, labeling, and validation to ensure that audio datasets support high-quality model training and evaluation.

Requirements

Five or more years of professional audio ML experience with deep expertise in digital signal processing and audio algorithm development is required.

Demonstrated ability to own end-to-end ML pipelines, guiding projects from initial experimental phases through deployment in cloud environments.

Strong proficiency in Python and deep learning frameworks such as PyTorch is essential for implementing and iterating on solution designs.

Capability to read and critically evaluate research papers, transforming theoretical concepts into production-ready code under real-world constraints.

Experience deploying ML systems for cloud-based inference, including containerization, orchestration, and performance optimization at scale.

A track record of influencing ML roadmaps, setting technical direction, and prioritizing research and infrastructure initiatives based on evidence.

Consistent skill in model quality assessment, balancing detailed user experience considerations with measurable business impact.

A stated degree is required as part of the educational background for this position.

Nice to have

A PhD or Masters in Computer Science or a related field strengthens candidacy for advanced research responsibilities.

Prior experience training large neural network models within academic or industry settings is valued for complex audio tasks.

Expertise in audio signal processing, including both classical methods and modern machine learning-based approaches, is preferred.

Publishing in top-tier conferences such as NeurIPS, ICML, ICLR, CVPR, and similar venues is recognized as evidence of research excellence.

Practical notes

The role requires a degree as stated and a minimum of at least five years of relevant professional experience.

Candidates must work onsite in San Francisco, as location is fixed for this position.

Typical interview steps

Data interviews commonly include a SQL or coding exercise, a statistics question, and a case study that mirrors real audio ML challenges.

You may be asked to design a metric, interpret an experiment, or build a small model to validate your technical approach.

Some employers use take-home analysis tasks to evaluate your ability to structure and execute an audio data project.

Expect detailed questions about past projects, with a focus on how your work influenced product decisions and handled trade-offs.

Interviewers often evaluate how you communicate uncertainty, manage trade-offs, and articulate business impact, not only the mathematical correctness of your solutions.

Bringing a clean write-up of a past analysis or experiment to the interview is well received and demonstrates thoroughness.

Good to know

Audio data R&D treats audio as a primary modality for AI systems, requiring specialized handling of time-series and spectral information.

Signal processing techniques and modern machine learning models power tools for data creation, curation, and evaluation in audio-centric workflows.

Teams combine research, engineering, and product functions to serve enterprise customers with demanding audio use cases.

Cloud platforms host production inference at scale, supporting high-throughput audio processing and low-latency responses.

Clear evaluation metrics link model performance to user outcomes and business impact, ensuring that technical progress translates into value.

Questions to ask

Good questions to ask the employer in the interview include what does success look like in the first six months, how is the team structured, and what is the current biggest challenge.

Asking about growth paths, the review process, and how audio ML initiatives are prioritized can provide clarity on long-term opportunities.

Asking about evaluation frameworks and how model improvements are measured demonstrates an interest in impact.

Employers expect questions, and well-prepared inquiries show that you understand the role and its context.

Career growth

Data careers grow toward senior analyst, staff data scientist, or data engineering lead roles over time.

Many professionals specialize further in machine learning, analytics, or infrastructure as they advance in their careers.

Cross-functional work with product and engineering teams becomes increasingly important at senior levels of responsibility.

The field evolves rapidly, so continuous learning and adaptation are integral parts of the job.

Professionals who can translate complex numbers and experiments into clear decisions tend to advance faster in their careers.

About the company

David Ai is hiring for Senior Machine Learning Research Engineer. San Francisco.