James Choi, Machine Learning Developer in Vancouver, BC, Canada
James Choi

Machine Learning Developer in Vancouver, BC, Canada

Member since August 27, 2018
James enjoys the challenges of extracting value from real-world datasets by wielding the latest machine learning and artificial intelligence tools in the Python stack. He has extensive experience in scientific research in academia and industry and thrives in collaborative teamwork.
James is now available for hire




Vancouver, BC, Canada



Preferred Environment

Python Stack, Atom/Vim, Git, AWS, Linux/macOS

The most amazing...

...thing I've worked on was applying my neuroscience expertise to develop new methods in the characterization of charge-stabilized nanostructures (nanobubbles).


  • Data Scientist

    2017 - PRESENT
    GN Science
    • Created a neural network-based text classifier and text summarizer with transfer learning (using Fast.ai and AWS SageMaker) for a client's proprietary data.
    • Built a data pipeline to scrape (ETL) more than 1 million PDFs from a website to model and rate a financial broker's integrity using the AWS platform (EC2, RDS, DynamoDB).
    • Implemented new clustering algorithms for the identification and characterization of neuronal cell types.
    • Developed and managed digital marketing campaigns (Facebook Ads, Google, AdWords Search/Display/Video), for big-budget (more than $100,000/month) clients.
    • Implemented A/B testing to optimize the customer journey through a conversion funnel.
    Technologies: Python, Fast.ai, MongoDB, PostgreSQL, AWS EC2, DynamoDB, Amazon SageMaker
  • Scientist

    2010 - 2013
    Revalesio Corporation
    • Set up electrophysiology laboratory facilities for in-house ion channel screening.
    • Established a semi-automated patch clamp pipeline for membrane-based target screening using proprietary cell lines.
    • Coordinated studies with leading institutions and CRAs to elucidate the biophysical mechanism of action for nanobubble therapeutics.
    • Actively contributed to the physical-chemistry characterization of nanoparticles in solution.
    • Devised new detection methods using nanopores.
    • Successfully detected and characterized cardiomyocyte activity modulation for A/B testing.
    Technologies: Single-cell Electrophysiology, Automated Patch Clamp, HEKA Patchmaster, MATLAB
  • Postdoctorate Fellow

    2006 - 2007
    Boston Children's Hospital | Harvard Medical School
    • Conducted single-cell-patch-clamp experiments on tissue slices from the mammalian visual thalamus.
    • Devised stereotactic injection techniques for the focalized injection of viral vectors.
    Technologies: Electrophysiology, Neuroscience, MATLAB, Igor


  • Current and Previous Projects in Data Science (Development)

    This is a blog that documents all of the projects that I have tackled thus far.


  • Paradigms

    Data Science, CRISP-DM
  • Other

    Software Development, Software Developer, Natural Language Processing (NLP), Supervised Learning, Machine Learning, Artificial Intelligence (AI), Clustering
  • Languages

    Python 3, Python 2, SQL
  • Libraries/APIs

    Pandas, NumPy, Scikit-learn, PySpark, Fast.ai
  • Tools

    GitHub, Git, Amazon SageMaker, AWS ECS, AWS Glue, AWS CLI
  • Platforms

    AWS EC2, Amazon Web Services (AWS), Jupyter Notebook
  • Storage

    MongoDB, AWS S3, PostgreSQL
  • Frameworks



  • PhD degree in Neuroscience
    1998 - 2005
    Brandeis University - Waltham, MA, USA
  • Bachelor of Arts degree in Biochemistry and Cognitive Science
    1993 - 1998
    Rice University - Houston, TX, USA
    International Language Academy of Canada Vancouver‎

To view more profiles

Join Toptal
I really like this profile
Share it with others