AI Speech & Audio Specialist

About the role

As an AI Speech & Audio Specialist at TELUS Digital, you will help advance AI systems that understand, process, and generate human speech. You will work with a global community to evaluate, create, and refine high‑quality speech and audio datasets that power next‑generation speech technologies.

Key responsibilities

• Evaluate AI‑generated speech (TTS) for naturalness, pronunciation, rhythm, tone, emotion and overall voice quality.
• Review and improve ASR outputs by identifying transcription errors, pronunciation issues, accents and linguistic nuances.
• Perform professional audio editing tasks such as segmentation, trimming, noise reduction, silence adjustment and quality enhancement.
• Analyze recordings with waveform and spectrogram tools to detect quality problems.
• Support creation and validation of speech datasets used for AI model training and evaluation.
• Conduct audio quality checks including loudness levels, background noise, clipping, distortion and export specifications.
• Contribute to voice AI projects including TTS, ASR, voice generation, voice similarity, voice mimicry and conversational AI evaluation.

Required profile

• Bachelor’s degree or equivalent experience in Audio Engineering, Sound Engineering, Music Technology, Linguistics, Phonetics, Speech‑Language Pathology, Computational Linguistics, Computer Science (Speech/AI focus) or related fields.
• Professional experience in audio production, sound design, dialogue editing, voice recording, podcast/audio post‑production, speech data creation or AI speech projects.
• Native‑level proficiency (C1/C2+) in the target language for language‑specific projects.

Required skills

• Proficiency with at least one professional audio tool (e.g., Adobe Audition, iZotope RX, Pro Tools, Reaper, Audacity, Logic Pro, Cubase, or other DAWs).
• Knowledge of audio editing and restoration, noise reduction, speech segmentation, spectrogram analysis, and audio quality assessment.
• Experience with voice/audio processing workflows.