About the role
You'll lead speech research for Ira, our upcoming voice assistant — TTS, ASR, and real-time multimodal voice for Indian languages and accents.
What you'll do
- —Research and train TTS/ASR and speech-to-speech models
- —Push quality on Indian languages, accents, and code-mixed speech
- —Build evals for naturalness, latency, and intelligibility
- —Collaborate with the multimodal (omni) model team
What we're looking for
- —Speech/audio ML experience (TTS, ASR, or speech LLMs)
- —Strong PyTorch and audio-data pipelines
- —Research depth in speech or adjacent areas
Nice to have
- —Indian-language speech datasets
- —Real-time/streaming audio
- —Voice cloning expertise
Sound like you?
We hire for skill over credentials. Tell us why you're a fit — links and projects welcome.
Apply for this role