🔧 Conventional backend system Vs ML backend system

One significant difference I've noticed in exploring machine learning in the backend compared to a traditional backend system can be likened to cooking a meal. In a conventional backend system, you have all the necessary ingredients, know the recipe, and have a clear idea of the expected result. In contrast, building an ML backend system is like having the ingredients and understanding the expected results, but not having a recipe.