Machine learning is hard, deploying to real business is harder

Marcin Druzkowski

Marcin will offer an example of how Ocado Technology is using data science to analyse and prioritize customer emails sent to a contact centre. He will share a number of useful tips on how to deploy a machine learning project into a real business environment, and what needs to be done to allow data science teams to work effectively within larger organisations.

Marcin will also outline tools such as Git, Docker, Jupyter and TensorFlow, and show how this technical stack can work to solve real business problems.

Required audience experience: A basic understanding of TensorFlow, cloud platforms and natural language processing is helpful.

Objective of the talk: Attendees can learn more about how they can put machine learning models into production and why software engineering principles like code versioning, continuous improvement and proper monitoring are crucial to the success of their models.

Keywords: TensorFlow, AI, machine learning, NLP, Google Cloud

You can view Marcin’s slides below:

Marcin Druzkowski: Deploying Machine Learning in Real Business

You can view Marcin’s presentation below:

Track 1
Location: Date: October 10, 2017 Time: 1:25 pm - 2:10 pm Marcin Druzkowski Marcin Druzkowski, Ocado Technology