J'ai postulé en ligne. Le processus a pris 5 semaines. J'ai passé un entretien chez The Smart Cube en juin 2022
Entretien
The took interview process was a month long.
Be aware they may take 5 rounds and reject you at last round.
1- The interviewer was very rude and didn't answered objectively, on asking clarifications, said" this so basic"
Questions d'entretien [1]
Question 1
Q- Why us?,
Q- Interview about hobbies, general areas, current market in Europe.
J'ai postulé via une autre source. Le processus a pris 2 semaines. J'ai passé un entretien chez The Smart Cube (Noida) en août 2022
Entretien
There were three rounds of interview, including one detailed case study round, during which candidates were evaluated on their analytical ability, problem-solving skills, and overall approach to real-world business scenarios
Questions d'entretien [1]
Question 1
Based on Case study
Past experience
case study during interview process need to provide solution on the spot
J'ai postulé en ligne. Le processus a pris 2 semaines. J'ai passé un entretien chez The Smart Cube en mai 2022
Entretien
I have 3+ years experience as a Data Scientist. Interview was not much difficult and interview in the first round mentioned that I mustn't give him defintions and textbook english, but answer in plain english and only give practical answeres.
According to a HR, company doesn't have a role called Data Scientist or Machine Learning Engineer.
Questions d'entretien [11]
Question 1
detailed discussion on my project on my resume with many counter questions during my explanation on why some algorithm was chosen instead of something else? why did you the choose the metric you did and how would that affect the business outcome? how have you decided that the approach you took is the right one to solve the business problem at hand?
Summary of SVM by summarising how it would work for regression outcome. (after my answer), what will you do if you have data that is not linearly seperable?
How will you clean text data? (question was specific on how to remove, i think it was extra dots, spaces, special characters or something else and it actually wasn't this broad and open-ended. )
what activates a neural network? (actual question was something else, I couldn't make anything out of it, so, after some dicussion this was the question)
Describe how you would monitor a deployed model on sagemaker? and during that monitoring, if you learn that model has started to give bad/incorrect predicitons, how you would trigger the training a new model and replace the existing one with the newly trained model without affecting the end-systems that are using it and those systems not noticing the shift?
