When evaluating the performance of a model, accuracy is usually the primary consideration. However, in business applications, speed is also a crucial factor. Even if you have the best model available, your users may switch to your competitors if they have to wait too long for the results. In this discussion, I will explain how to balance accuracy and speed to achieve the best value for your business. 

During the talk, you will learn how to determine which parameter is the most important in your case, identify the bottlenecks that may affect the performance of your model, and find solutions to optimise your model according to your needs. I will provide practical examples of how to optimise a model, using technologies such as Pandas, XGBoost, LGBM, cProfile, and yappi.