SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
David AI is an audio data research company founded in 2024 by former Scale AI engineers. The company has raised $50M in Series B funding from top-tier investors including NVIDIA, Meritech, and First Round Capital, and counts most FAANG companies and leading AI labs as customers.
The Research team sits at the intersection of cutting-edge research and production systems, transforming raw audio into high-signal datasets for AI labs and enterprises. The team owns the full ML lifecycle—from researching novel speech processing algorithms to deploying models that process terabytes of audio daily.
As a Research Engineer, you will build algorithms and systems that turn large-scale speech and audio collections into high-quality datasets. You'll tackle core questions: What makes a recording useful? How do we measure and improve dataset quality?
Key responsibilities:
- Design algorithms and metrics using advanced signal processing and cutting-edge ML models to derive insights from thousands of hours of audio data.
- Build scalable data processing pipelines for analyzing, filtering, deduplicating, and curating large speech and audio datasets.
- Develop production-grade inference algorithms, pipelines, and APIs with cross-functional teams to unlock key insights for customers.
- Collaborate with the Operations team to gather useful training and evaluation datasets to improve model quality.
- Architect systems that enable resilient, durable inference and evaluations.
Requirements:
- 3+ years of professional engineering or applied research experience in data systems, algorithms, applied ML, information retrieval, or speech and audio.
- Strong Python skills and experience building maintainable, production-quality software.
- End-to-end ownership of ML pipelines, from proof-of-concept to production deployment.
- Ability to translate research papers and ideas into high-quality, production-ready code.
- Experience deploying ML systems for production inference with cloud technologies.
- Ability to assess model quality in the context of user experience and business value.
Bonus qualifications:
- Experience with speech, audio, or digital signal processing.
- Experience with data curation, search and retrieval, embeddings, clustering, deduplication, or sampling.
- Experience building model inference systems or integrating pretrained models into production workflows.
- Publications in top conferences (NeurIPS, ICML, ICLR, CVPR, etc.).
- Open-source software contributions or maintenance experience.
- PhD or Master's in Computer Science or related field.
