Design patterns for continuous intelligence
We show how the deployment design pattern for continuous intelligence systems differs from those used in traditional ML.
- continuous-intelligence
I completed my PhD in physics at the University of New South Wales in 2010. Subsequently, I have applied data science and machine learning in a range of industries, including financial services, media, telecommunications, and retail.
We show how the deployment design pattern for continuous intelligence systems differs from those used in traditional ML.
MLOps is a rapidly growing field aimed at standardising and streamlining the lifecycle of ML models, from development and deployment through to ongoing maintenance.
The ability to test decision-making algorithms without actually deploying them to live systems is an important feature for automated decision-making systems to have.
We examine how experimentation platforms used by search engine and social media companies have been designed to handle a range of experimentation issues that occur at scale.
Clear and unbiased analysis methods are critical to understanding the experiments that businesses run in-market on real users.
Uplift models are used to identify customers who are most likely to respond positively as a result of receiving some intervention.
What is a contextual bandit? How does it enable continuous intelligence and real-time decision-making?
Fairness and understanding hidden biases in algorithms used for decision making are increasingly important.
A/B testing is a simple way for companies to make design decisions under uncertainty.
If you torture the data long enough, Nature will confess.