The overall process took them 45 days! I mean like really? I really should have moved on. Round 1: 1. Home assessment. A stock dataset was given. Clustering and decision tree problems specifically mentioned to be implemented. This is where I put alot of efforts in building models, implementing grid search, making sure I mention evaluation metrics, tree plot, evaluation metrics for clustering was also sent to them. Round 2: 1. How would you tokenize sentence without using spacy or nltk. Using countplot and re.split() ofcourse. 2. A dataset was shown on screen. It was a classification problem. Just try to explain what steps you would take to make ml model. The questions here were too simple. Overall thoughts: 1. No matter how good you are, they will probably look at the number of years of experience you have. 2. They will waste your time by taking multiple rounds of you only to hire some one experienced. 2. Because I don't know where exactly did they measure my problem solving capabilities.
