From the course: Transforming Supply Chains with AI: A Leader's Road Map

Unlock this course with a free trial

Join today to access over 24,800 courses taught by industry experts.

Embracing Agile

Embracing Agile

From the course: Transforming Supply Chains with AI: A Leader's Road Map

Start my 1-month free trial Buy for my team

Embracing Agile

“

- AI implementation is a lot different from traditional software deployment where we have clear go/no-go decision points. Firstly, AI involves significant ambiguity. Until we see the first cut of an AI model in action, when it's exposed to a real-world scenario, we cannot tell with certainty how well it will perform with the unseen data. For example, you can train a demand forecasting AI model with as much data as possible and it might perform with high accuracy in a test environment. But there might be additional hidden patterns in the real-world data that need some more tweaking or simply retraining of the algorithm. Secondly, in theory, AI follows a virtuous cycle of development which effectively means once an AI product is used, it generates data. This data then gets fed into the algorithm, therefore improving its performance and hence increasing usage, which then generates more data. And the loop continues. This continuous improvement of AI requires supply chain leaders to…

Contents