Watch a Data Scientist Ace the DoorDash Case Study Interview Question: Minimize Wrong Orders!

[@rohitverma1057]

Do we really need a machine learning system for this ? I mean it's in the hands of user if he places wrong order on system. You may recommend him any nice thing. But he may still add wrong item in cart. Also, simple double verification of address feature should help in wrong address selection. Tbh NO machine learning is required here. Some engineering level app features.

[@ShubhamDash14]

How does this solution solve anything? I would rather design a vision / text based solution to verify the order from the dasher's side. The idea would be does the order match the description of the order. Can use siamese network to measure similarity between multimodal embeddings

[@rishankijain]

I think she was struggling a lot. Very incoherent ...Zarrar was trying hard to give cues.

[@jordanberninger9590]

yea lol i def wouldnt say she "aced" this

[@ajayloves]

i don’t think this will work…. wrong order is only when customers raise the request this is wrong order. you need good data and features to design the model