Deploying Machine Learning Models

by Carl Stanley

Thursday 22 Apr 2021, 15:00 → 16:00 UTC

Zoom connection details: https://cern.zoom.us/j/98750947196?pwd=N1RTS3EzazBha0tURVJOcEZKREFxZz09 Meeting ID: 987 5094 7196 Passcode: 613113 (Zoom)

Data Science is one of the most popular job titles of the 21^st Century, with huge claims on the impact that it’s making across the entire business world. But how does what essentially boils down to a (sometimes quite simple) linear algebra problem actually make all of these breakthroughs, and more importantly make all of these people so sought-after in the job market? I’ll walk through the business-oriented data science workflow, and how we can go from experimentation with machine learning (just with customer data instead of particle interaction data), all the way through to deploying models to serve predictions at scale, and actually bringing about business value for our stakeholders.

 

About the speaker:

I graduated from Queen Mary in 2017 with a Masters in Physics, and went on to the BT graduate scheme as a business analyst. I then became one of the first data scientists sitting directly in BT’s B2B (business to business) arm, and became heavily responsible for defining the machine learning strategy for our space. I moved on to lead a 20 person team that would go on a 12 month journey from 0 to hero in the machine learning space, where we now have a live suite of models helping to optimise our incident workflow daily.

I’m currently working on a recommendation engine for our consumer space, where I’m aiming to provide personal recommendations on our mobile products (EE) to our broadband customer base.

Slides Deploying_Machine_Learning_Models_-_high_level.pptx.pdf