Menu open

You will work on the core of our products, the Natural Language Processing & Generation components. You will develop, implement, evaluate and document user-centered systems with Machine Learning / Deep Learning methodologies. Furthermore, you may integrate, test, and do research on software components that you helped develop or were shipped by our partners.  

What you will be doing

  • Work on all parts of the NLP & NLG pipeline, incl. data preparation, learning & inference for speech-to-text, intent recognition, named entity recognition and dialogue engineering
  • Develop requirements and solutions for novel voice user interfaces
  • Deliver excellent, production-ready code to our system following coding principles, such as clean code (incl. test-driven development, thorough documentation)
  • Deploy in Cloud (AWS), Docker containers in a Microservice environment

Requirements

  • Minimum Bachelor’s degree in Computer Linguistics, Computer Science, Mathematics or similar
  • Demonstrative experience in training ASR models based on traditional hybrid and/or recent end-to-end (Deep Learning) approaches. Would be ideal if the candidate has experience in working with Kaldi
  • Strong familiarity with natural language processing techniques and experience with training NLP models
  • Solid understanding of dialogue engine and has experience in preparing chatbots, preferably using Rasa.
  • Excellent Python skills and experience in frameworks such as TensorFlow, PyTorch, Keras, pandas
  • Excellent written and oral communication skills in Bangla & English
  • Experience in Agile/Scrum project methodologies.
  • High responsibility and team-oriented work attitude in an agile
    environment

Good to Have

  • Knowledge in the following development technologies will be appreciated:
    • Audio signal processing
    • Deploying Docker containers for Microservice environments
    • Linux
    • Versioning (Git)
  • Additional languages (e.g. Japanese)
  • A good reading & comprehension ability pertaining to research papers & patent documents
Back To Top