I am generally interested in the area of representation learning. More specifically, I am interested in the following areas: semi-supervision, self-supervision, and model robustness. At Carted, I am currently working on representation learning from raw HTML web pages as well as various use-cases around NLP.

I enjoy implementing research ideas, sometimes incorporating them in practical applications, and communicating my implementation details through articles. I advocate for highly readable and self-contained code. I primarily use Python for programming. I am proficient in TensorFlow and a beginner in PyTorch.

I have a flair for open-source initiatives. In particular, I have made contributions to Keras, 🤗 Transformers, TensorFlow Hub (highlights: DeiT, ConvNeXt, Vision Transformers, MLP-Mixers, CartoonGAN, DeepLabV3), Keras Examples, TensorFlow Addons, and Neural Structured Learning. Off the work, I like writing technical articles, working on applied Machine Learning ideas, and giving talks at developer meetups and conferences.

For my community contributions and innovative projects, the Intel Software Innovator Program recognized me as one of their top innovators in 2019. For my open-source contributions, I received the Google Open Source Peer Bonus Award in 2020 and 2021, TensorFlow Top Contributor Award in 2021.

GitHub Twitter LinkedIn Gmail

Timeline:

  • Machine Learning Engineer, Carted (June 2021 - Present)
  • Deep Learning Associate, PyImageSearch (June 2019 - June 2021)
  • Data Science Instructor, DataCamp (August 2018 - June 2019) (on contract)
  • Software Engineer, TCS Research and Innovation (January 2018 - August 2018)
  • Software Engineer, Tata Consultancy Services Limited (July 2017 - January 2018)
  • Intern, CareerIn (Dec, 2016 - Feb, 2017)

Badges I proudly endorse:

An honour to be their son 🙂

  • Tapas Kumar Paul
  • Baby Paul