Senior Researcher - Text to Speech | Bengaluru

Smallest; Bengaluru; On-site; INR 4,50,000 - 6,75,000; Full time

Job summary

A leading technology firm in Bengaluru is seeking a skilled individual to lead research on Text-to-Speech models. The ideal candidate will have a strong background in speech generation, experience with deep learning frameworks like PyTorch, and knowledge of real-time TTS systems. A Master's or PhD in a relevant field is required. Join us to design and improve innovative TTS systems catering to diverse accents and styles while collaborating with expert teams.

Qualifications

• 3–6 years of specialized experience in speech through academia or industry.

Responsibilities

• Lead research on Text-to-Speech models focused on naturalness and expressiveness.
• Design and train TTS systems for real-world voices across accents.
• Improve streaming and low-latency speech synthesis pipelines.

Skills

Text-to-Speech / speech generation research

Deep learning frameworks (PyTorch preferred)

Real-time or low-latency TTS systems

Modern TTS architectures

End-to-end thinking from data to deployment

Multilingual, expressive, or accented speech synthesis

Education

Master’s or PhD in Speech, ML, or a related field

Job description

What you’ll do

• Lead research on Text-to-Speech models focused on naturalness, expressiveness, latency, and robustness

• Design and train TTS systems for real-world voices across accents, languages, and speaking styles

• Improve streaming and low-latency speech synthesis pipelines

• Experiment with architectures, loss functions, and data strategies (multi-speaker training, style modeling, distillation, data augmentation)

• Translate research ideas into production-ready TTS systems

• Collaborate closely with infra, product, and voice engineering teams

What we’re looking for

• Strong background in Text-to-Speech / speech generation research

• Hands-on experience with deep learning frameworks (PyTorch preferred)

• Experience with real-time or low-latency TTS systems

• Familiarity with modern TTS architectures (Tacotron-style, FastSpeech, VITS, diffusion-based, neural vocoders)

• Ability to think end-to-end: data → model → inference → deployment

• Prior work in multilingual, expressive, or accented speech synthesis is a strong plus

Great to have

• Publications in top speech / ML conferences

• Experience deploying TTS models in real-time production

• Exposure to conversational AI or voice agents

Years of Experience

• 3–6 years of specialized experience in speech through academia or industry

Education

• Master’s or PhD in Speech, ML, or a related field

Note: We often make exceptions and hire brilliant candidates regardless of years of experience or education. Proof of work is paramount.