Scott Graham
Data Scientist - Developer - Sucker for the Shiny and New



ML-Agent : Machine learning model server for on-demand predictions written in Python using a Flask micro-framework.

ML-Toolbox : Docker image to simplify experimentation with common machine learning tools through Jupyter Notebook.

ML-Dashboard : Docker image to simplify deployment of ML visualization dashboards.

AudioBot : Upload and play audio clips on demand through a raspberry-pi + speaker.

Emojibot : Slack Bot with Tensorflow LSTM model to predict appropriate reactions to add to text.


Vowpal Wabbit : Machine learning library implemented in C++, for fast online or batch learning using a wide array of techniques.

TensorFlow - SKFlow : Sklearn interface to popular TensorFlow tool from Google for large-scale machine learning.


I am currently working as an Sr. Optimization Engineer at Nanigans implementing machine learning algorithms for real-time bidding and advertising campaign automation.I love designing technical solutions using machine learning and bringing them to life to improve product performance and usability. I'm passionate about code quality and continuous improvement but always try to balance practicality and development speed.

I enjoy learning new machine learning topics, technical tools and programming languages. I feely strongly about supporting OSS projects and contribute where I can. In between being a husband, home-owner, and father of 2, I build bots (both the physical and virtual varieties), soak up as many books and podcasts as possible, and love to travel, cook, and get out into nature for hiking, camping and skiing.