Candidatei-me online. O processo levou 2 semanas. Fui entrevistado pela Monedo em mai. de 2015
Entrevista
I've applied online and the same day got an invitation to take a technical test. The test was scheduled for 4 hours and I decided to take it over the weekend. The task was a binary classification problem, with testing and validation datasets provided, to be completed in R along with short writeup with observations. I use R on daily basis in financial sector (e.g. rating/scoring model development) so I'm familiar with tools and industry techniques used.
For the test solution I have used gbm (generalized boosted regression models), with post-statistics like ROC, and mentioned other applicable methods to apply in such binary case (e.g. widely used in financial/banking - which the company allegedly is operating in - logistic regression).
After almost 2 weeks I've got a reply from HR saying the CDSO's feedback is negative without actually providing the said feedback at all. I was trying to get more detailed feedback from the company with no success however. Sadly this is my second interaction with the same company as several months back I've send my CV for general consideration and possible fit into Data Science team. I've had an initial chat with an HR and absolute no response after (despite my several requests on my application status update).
Seems like the start-up company profile needs to be more seasoned in order to treat non-entry level candidates with proper attention. Overall I don't recommend, at least not in current shape.
Perguntas de entrevista [1]
Pergunta 1
binary classification problem using testing and training sample
Candidatei-me online. O processo levou 1 dia. Fui entrevistado pela Monedo (Hamburgo) em fev. de 2018
Entrevista
Given an anonymised data set and asked to build a predictive model in R. Had to take a morning off work to do it. Zero feedback; ghosted by the company afterwards. I suspect they are just looking for people to analyse their data for free.
Perguntas de entrevista [1]
Pergunta 1
Data set contained a variable that was a perfect predictor on the training set but with no predictive power in the validation set.
Candidatei-me online. O processo levou 2 meses. Fui entrevistado pela Monedo em nov. de 2017
Entrevista
4 steps with a very long waiting period in between them:
- Coding test (4h): they give you a very abstract train dataset and a test dataset, and you have to perform a classification in R. You must write a report on your work.
- Phone call (15min): a call with the RH to talk a bit about you, your cv, experiences etc...
- Technical interview (phone call, 1h or 30min, don't remember quite well): a call with the person in the data science team who reviewed your test. Basic questions about the algorithms used, see if you understand how the algorithms work, on which kind of data, etc...
- On-site interview with oral presentation (4h I think): they give you a date and a time to finally come and meet them. They ask you to prepare a presentation of 20min about something related to data science. I didn't go as I found another job in between, so I don't know more.
I applied on the 07/10, and was supposed to come to the on-site interview on the 15/12. More than 2 months for the whole process.
Perguntas de entrevista [1]
Pergunta 1
What is a classification tree and a random forest? Why is the random forest more robust than the tree?
O processo levou 6 semanas. Fui entrevistado pela Monedo em mai. de 2017
Entrevista
After one month of my application, I got the test from HR department. It was about the binary outcome prediction. They provided training set and testing set, and expect you to predict the outcome in testing set by training set. I used random forest and SVM to predict.
Unfortunately, I was rejected, and I didnt get any response according to the report. Just a rejection letter without any specific reason. I think that is normal for this company.