[Hiring] Research Scientist, Speech & Audio @Innodata Inc.

Role Description

Innodata is hiring a Research Scientist to own the science behind building data and evaluations for speech and audio models. You will partner directly with customers and frontier labs to advance ASR, text-to-speech, speech-to-speech, and conversational voice models.

• Define how Innodata designs, structures, and evaluates audio data for speech and audio models.
• Translate model requirements into concrete data specifications: modalities, transcription and annotation schemas, sampling, and evaluation criteria.
• Build evaluation methodology that goes beyond word error rate, focusing on:
  • Semantic accuracy
  • Robustness to noise and accent
  • Code-switching
  • Diarization error rate (DER)
  • Naturalness and intelligibility of generated speech
  • Streaming latency
• Decide how audio should be structured, enriched, and sampled for model objectives across various conditions.
• Partner with the transcription and linguistics lead to create transcription specifications.
• Collaborate with the audio solutions and engineering team to ensure audio collection meets model objectives.
• Run experiments to validate data decisions and their impact on model performance.
• Design adversarial evaluations to identify weaknesses in speech systems.
• Publish findings to advance the field and build trust with customers and labs.
• Work with teams to operationalize specifications into plans.

Qualifications

• 5+ years of hands-on industry experience in speech or audio ML.
• Bachelor's degree in computer science, electrical engineering, or a related field; advanced degree preferred.
• Experience training and evaluating speech or audio models with strong PyTorch fundamentals.
• Fluency in toolchains and metrics used in speech work: ESPnet, NeMo, SpeechBrain, Kaldi, HuggingFace.
• Hands-on experience with multilingual, accented, dialectal, low-resource, or code-switched speech.
• Ability to think critically about datasets and evaluation sets.
• Track record recognized in the field: first-author publications or strong open-source contributions.
• Ability to communicate clearly with both expert and non-expert audiences.
• Bonus: Interest or experience in responsible-AI evaluation and red-teaming.

Requirements

• Strong understanding of evaluation methodologies and metrics.
• Experience with audio data collection and processing.
• Ability to run and analyze experiments rigorously.
• Excellent communication skills for collaboration and documentation.

Benefits

• Competitive salary range: $160,000 - $185,000 per year.
• Opportunities for professional development and growth.
• Collaborative work environment with leading experts in the field.