about 8 hours ago
Sandy, UT, USASenior / Mid Level
H1B Sponsor
Responsibilities
- Design and maintain a speech-oriented test suite covering quality, cost, and latency across multiple languages.
- Monitor the industry for new state-of-the-art transcription and speech-to-speech models.
- Design and evaluate techniques to optimize speech model usage for operational deployment.
- Produce clear, quantitative evaluation reports and model recommendations.
- Contribute to the broader model evaluation framework maintained by the NLR team.
- Stay informed of advances in speech AI technologies.
Requirements
- MS in computer science, electrical engineering, computational linguistics, or a related field focused on speech or audio processing.
- Three or more years of hands-on experience with speech AI systems, including ASR, TTS, or speech-to-speech models.
- Experience designing evaluation methodologies or test suites for AI systems.
- Strong quantitative and analytical skills with experience producing rigorous benchmark results.
- Proficiency in Python and familiarity with speech processing libraries and tools.
- Experience with cloud-based infrastructure (AWS, Azure, or GCP).
Benefits
- Flexible hybrid work model with 2 days in the office and 3 days remote each week.
- Opportunities for internal career growth across multiple roles and disciplines.
- Collaborative and creative work environment in a market-disrupting company.
Tech Stack
Categories
AI & MLData Science