Responsibilities

  • Delivering superior working solutions to speech recognition and speech audio processing problems arising in Uniphore’s products and services
  • Improvement of existing algorithms and features of Uniphore’s products and features with principled and scalable methodology
  • Annotation, collection, clean-up and/or testing and validation of speech data Defining testing norms and procedures for speech systems
  • Contributing to Uniphore’s IP

Required Qualification

Technical/Job Skills

  • Operating Systems: Windows & Linux. Languages: Python (proven mastery required), Java
  • Proven mastery of, and track record in all aspects of LVCSR
  • Thorough command of, and demonstrable experience in modern machine learning including deep learning and sequence-to-sequence modelling as applicable to speech recognition and speech signal processing
  • Experience with speech recognition engines and toolkits (Kaldi, and one or more of Sphinx, Google Speech Alexa, IBM Watson & Nuance)
  • Should be at ease with both statistical language modeling and grammars
  • Good experience with (and knowledge of) statistical signal processing
  • Extensive exposure to, and experience with speech data, preferably in multiple languages
  • Solid basics in statistics and linear algebra
  • A scrupulous habit of documenting work
  • Keeping up with developments and progress in ASR and Machine Learning

Functional Skills 

  • Ability to handle multiple projects
  • Passion and commitment to work in a start-up environment

Desirable Qualifications 

  • Proven practical experience with modern speech recognition engine technology, such as DNNs, variant neural methods and WFSTs
  • Good knowledge of (and experience with) phonology, phone sets, pronunciation dictionaries and G2P
  • Experience with collecting and training models with acoustic data and text data Experience of team leadership, line management and mentoring
  • Project management experience (preferably with Jira)
  • Knowledge of (and experience with) NLP problems and techniques
  • Software development experience with a history of shipping products
  • Experience working with Indian languages (data collection, annotation, Romanization, ASR/TTS systems, NLP)
  • Connectedness in the landscape of speech technology in India

Apply for position

  • Accepted file types: doc, docx, xls, pdf.
    .doc,.docx,.XLS,.PDF
  • This field is for validation purposes and should be left unchanged.