ML Engineer - Speech

Zoho; Dadri; On-site; INR 1,200,000 - 2,400,000; Full time

Job summary

Zoho is seeking an ML engineer specialized in speech to fine-tune models on real call audio and build scalable training/evaluation pipelines. You will handle multilingual and code-mixed data, targeting production-grade accuracy and low latency.

You will collaborate with linguists on text normalization and pronunciation, evaluate architectures, and move successful models to production with the platform team.

Qualifications

• 3–4 years in ML with at least 18 months in speech/audio.
• Strong Python and PyTorch; able to read a paper and implement it.
• Hands-on experience fine-tuning at least one production speech model.
• Solid grasp of speech fundamentals—mel-spectrograms, acoustic models, vocoders, encoders/decoders, evaluation.

Responsibilities

• Fine-tune and adapt open-source speech models on our proprietary call audio.
• Build training and evaluation pipelines for multilingual and code-mixed speech.
• Own model quality against metrics that matter operationally—entity and numeric accuracy, latency to first response.
• Design and run data curation at scale: pseudo-labelling, speech enhancement, quality filtering on messy real-world audio.
• Work with our linguist on text normalisation and pronunciation handling.
• Evaluate candidate architectures, make the call with evidence, and ship the result to production with the platform team.

Skills

Python

PyTorch

Machine Learning

Speech processing

Model fine-tuning

Multilingual speech

Tools

NeMo

ESPnet

SpeechBrain

Coqui

Job description

• Fine-tune and adapt open-source speech models on our proprietary call audio
  
• Build training and evaluation pipelines for multilingual and code-mixed speech
  
• Own model quality against metrics that matter operationally — entity and numeric accuracy, latency to first response — not just aggregate error rates
  
• Design and run data curation at scale: pseudo-labelling, speech enhancement, quality filtering on messy real-world audio
  
• Work with our linguist on text normalisation and pronunciation handling
  
• Evaluate candidate architectures, make the call with evidence, and ship the result to production with the platform team
  

Requirements

Must have skills:

• 3–4 years in ML, with at least 18 months on speech or audio specifically
  
• Strong Python and PyTorch; comfortable reading a paper and implementing it
  
• Hands-on experience fine-tuning at least one production speech model
  
• Solid grasp of speech fundamentals — mel-spectrograms, acoustic models and vocoders, encoder-decoder vs transducer architectures, evaluation methodology, sampling rates and what they cost you
  
• Understanding of how modern speech systems are actually built: self-supervised encoders, neural audio codecs, LM-based generation, flow matching
  
• Experience with genuinely messy audio, not only clean benchmark datasets
  

Nice to have:

• NeMo, ESPnet, SpeechBrain, or Coqui
  
• Telephony-band or contact-centre audio
  
• Multilingual or code-switched speech work
  
• LoRA/PEFT, distributed training
  
• Open-source contributions or publications in speech