Senior Data Scientist, Data Flywheel
Core
Characterize conversational audio data, design active-learning loops, and build data strategies to improve speech models at scale.
Role type
Senior Data Scientist (Data Flywheel & Active Learning)
Builds
Curated datasets, active-learning loops, and model-improvement pipelines for voice AI
Domain
Voice AI / Speech-to-Text (STT) / Audio Data Science
Deliverable
production ML models
Required skills
Python, data characterization, active learning, data selection, model debugging, automation, data pipeline engineering
Preferred skills
ASR/TTS experience, multilingual/code-switched data handling, ensemble labeling, pseudo-labeling, LLM-assisted annotation, data provenance management
Technologies
Python, data tooling, speech/audio models, NLP models
Responsibilities
Understand and characterize conversational data (languages, conditions, domains, speakers, quality); Design and build active-learning loops to prioritize high-impact work; Design workflows and tooling for human-in-the-loop annotation; Build curated datasets and benchmarking methodologies; Run experiments to validate or overturn data strategies; Automate model adaptation into repeatable pipelines
Seniority
Senior, hands-on IC
