System Design · Snapchat · Hard
In an interview setting, respond to the following two machine learning system design prompts. First, choose the end-to-end ML system you understand most deeply and describe it in full. Cover how the problem is framed, where the data originates, how features are constructed, what the offline training process looks like, how online inference is performed, how the model is refreshed, how the system is monitored, and what trade-offs exist between the offline and online portions.…
Checking your access…