Train the models powering the next generation of Voice AI.

A fast-growing AI technology team is looking for a hands-on Voice AI Engineer who has already worked on training or pre-training foundation models .

This is not a conventional ML application role. You’ll work at the model layer across speech, foundation models, large-scale datasets and production optimization .

ABOUT THE OPPORTUNITY

You’ll own the journey from speech/text data → model training → evaluation → optimization → production deployment , solving challenging problems across accuracy, latency and inference efficiency.

The role is ideal for an engineer with strong Speech AI / ML / NLP fundamentals who has actually trained models rather than only consuming existing AI APIs.

WHAT YOU'LL OWN

- Develop, train, fine-tune and evaluate foundation models for Voice AI .
- Build speech solutions across ASR, TTS, speaker identification and voice activity detection .
- Work on conversational AI and speech-language models.
- Prepare and optimize large-scale speech and text datasets .
- Design training pipelines and run model experiments.
- Work with GPU-based and distributed training workloads.
- Improve model accuracy, latency, robustness and inference efficiency.
- Apply model optimization techniques including quantization .
- Debug model, data and training-pipeline issues.
- Support production deployment and performance optimization.

THE IDEAL CANDIDATE We're looking for someone who has actually trained or pre-trained foundation models ,



not someone whose experience is limited to prompting, API integration or using pre-trained models.

MUST HAVE

- 3–4 years of hands-on experience in ML, Speech AI, NLP or closely related areas.
- Demonstrated experience training/pre-training foundation models .
- Strong understanding of Transformers and deep-learning architectures .
- Hands-on experience with ASR and/or TTS models .
- Strong Python skills.
- Hands-on experience with PyTorch and/or TensorFlow .
- Experience with distributed training and GPU workloads.
- Experience handling large-scale speech/text data pipelines.
- Model evaluation, optimization and production deployment exposure.
- Strong analytical and debugging skills.

VALUABLE TO HAVE

- Multilingual or Indian-language speech models .
- Whisper, wav2vec 2.0, HuBERT, NeMo, SpeechBrain or Hugging Face.
- Audio preprocessing, augmentation or annotation.
- Dataset quality improvement experience.

WHY THIS ROLE

- Work directly on Voice AI foundation-model development .
- Go beyond application-layer GenAI into actual model training .
- Work with large-scale speech datasets and GPU infrastructure.
- Solve challenging problems around accuracy, latency and inference .
- Build expertise across ASR, TTS and conversational AI .

R OLE DETAILS Role: Voice AI Engineer / Speech ML Engineer

Experience: 3–5 Years

Domain: Voice AI / Speech ML / NLP

Employment: Full-Time

Location: HSR Layout Bangalore

📌 Voice AI Engineer | Foundation Models, ASR & TTS | 3–5 Years (Bengaluru)
🏢 TalixoHR
📍 Bengaluru