I applied online. I interviewed at dunnhumby (London, England) in Apr 2024
Interview
Interview consisted of multiple rounds. 1st screening call, 2nd 1 hour technical interview with senior scientist, 3rd 1 hour online MCQ test testing data science knowledge and 4th case study based interview.
Overall experience was negative. Every interviewer I interacted with seemed either distracted or disinterested in talking from the start. Even basic pleasantries were exchanged at the start of any of the interviews.
Breakdown by parts:
1: Technical Interview: questions about previous research experience and a surprise case study question (which I was not told would occour by the HR)
2: online MCQ test on hackerearth consisting of 29 numerical ability questions, most questions designed to trip you up with subtle differences in the options or were GRE level math questions on linear algebra.
3: Asked for a rather detailed case study based on very vague outline. Given good time to prepare, however nothing was asked based on the work I did in the case study in the interview. Two questions were asked based on K-means and recsys, because I had mentioned those points in the case study — but the questions did not relate to the work in the case study at all. That is to say doing the case study tested nothing as the questions asked on it could have been done in isolation to measure the same effect. Was then asked several critical thinking type questions based on time series data, estimation and graph comprehension. One of the Interviewer was inarticulate and unclear in asking several questions with little context provided. Then there was a coding test, which asked easy coding questions like finding median of an array, length of array with non-recurring numbers and permutation generation, followed by some high school python questions to predict an output.
I was told at the end of the interview that I would be given feedback regardless of the outcome. No such feedback has been provided despite a mail to the HR following a generic rejection letter. If a candidate has spent several hours in your interview pipeline with several extra days spent in doing a case study for the interview, the least that can be done is to provide feedback on the task performance. Disappointing to say the least.
Interview questions [1]
Question 1
[technical screen questions]:
1. How would you design a system to pick which cereals to stock in a new upcoming Tesco
[Post Task]
1. Describe k-means, how does the algorithm update? What distance metrics can you choose? How to select k?
2. Describe how a recommender system can be used to recommend products.
3. Design a system to predict customer churn in next 4 weeks. Past data is described in a graph.
4. Estimate the number of Tesco customers who buy 2 pints of milk in a month within the UK.
5. Draw the distribution curve, mark the mean, median and mode.
[coding questions]
1. Given an array, return the mean and median, cannot use inbuilt functions.
2. Return all paired permutations of an array except self.
3. Return length of array consisting of non-repeating elements. Given array consists of repeated elements.
Followed by some high school level python output prediction questions. Quite insulting that something like this is even asked at this stage to be honest.
interview was smooth and very clear steps. they keep you updated on each steps. They is a official steps to hire . After you submit your application, the recruitment team reviews your CV, experience, and motivation to assess whether your background aligns with the role.
I applied through a recruiter. The process took 4 weeks. I interviewed at dunnhumby (London, England)
Interview
1. A 30-minute online interview with the hiring manager.
2. A Python programming test conducted at the office. No laptop was provided, and I was asked to write codes on paper.
3. A case study and presentation at the office. No data was provided for the task. The interviewers were insight managers who appeared to have little knowledge of data science. The questions focused mainly on insights and marketing strategies, which are not typical duties of a data scientist and were not mentioned in the job description. This made the job post very misleading. Additionally, the feedback provided was unreasonable and disrespectful, given that what they were looking for was not within the usual scope of a data scientist’s role.
Interview questions [1]
Question 1
How do you think the company should handle customers like this?
I applied online. The process took 4 weeks. I interviewed at dunnhumby (Dublin, Dublin) in Jul 2024
Interview
Typical interview process and questions. Two stages, one virtual interview followed by on site interview. Both interviews contained technical and behavioural sections, discussing prior experience in detail. Discussed details about the role, team, salary and compensation etc. Farily standard.
Interview questions [1]
Question 1
Discussed previous technical projects and experience