Ho presentato la mia candidatura tramite un selezionatore. La procedura ha richiesto 4 settimane. Ho sostenuto un colloquio presso CVS Health (Hartford, CT) nel mese di set 2023
Colloquio
Let me just tell you, that was hands down one of the worst interview experiences I’ve ever had! I mean, the SQL questions were ridiculously tough, and don’t even get me started on the corderpad environment. It had so many issues, I couldn’t even run my CTEs! Even the interviewer tried solving and gave it a shot and just gave up because even he couldn't resolve technical issues. And the interviewer seemed like he had this one solution stuck in his head and was expecting me to read his mind or something. There was no room for creativity or different approaches. But wait, it gets even worse. The Python questions were a whole other level of tough. I was handed a complete ML project and was expected to debug and find issues in the code. I mean, who does that in an interview?! Even faang interviews would not be that tough. Debugging an entire pipeline code isn’t something you just throw at someone during an interview. He literally asked me to debug scikit learn pipeline and all. Looking at 100 lines of code itself will overwhelm a candidate. No one would implement ML code from scratch at a company. How do you expect a candidate to understand and debug on the spot. It was like they were setting me up to fail from the start! Honestly, it was a frustrating and disappointing experience, far from what I’d expect from any professional setting. I wonder if they treat all their candidates like this, or if I just got the short end of the stick. Just test sql skills and python dsa. Test the character. Once anyone joins a company, they google to implement ML. No one develops anything from scratch. Easily the most disappointing interviewer I've ever come across. I feel happy for not selecting because I don't need to work alongside with him.
Domande di colloquio [1]
Domanda 1
- 3 sql questions: Extremely tough. You need to use every sql function out there to answer. Yes. You need to use cte, window functions, aggregations to solve them. Ridiculous right.?! - 3 python questions: Impossible to answer. I am given a whole ML project and asked to debug on spot. It is 100+ lines of code and expected to identify bugs in pipeline. You are expected to remember every data science package. Scikit learn pipeline etc., Even faang companies will not ask you this tough. Clearly the interviewer wanted to showcase his selection skills to higher management. He flaunted his skills than reviewing candidate skills.
The phone screen was surprisingly in-depth, lasting about 40 minutes and focusing heavily on real-world applications of data science. We discussed model development, particularly for predicting hospital readmissions, which was quite challenging. I had just practiced a similar scenario on PracHub that mirrored this section almost flawlessly. Following this, I went through a technical round where I tackled questions on feature engineering and demographic fairness. Ultimately, I received an offer but decided to decline it after weighing my options.
Domande di colloquio [1]
Domanda 1
How would you build a model to predict 30-day hospital readmission risk for chronic patients? Walk through your feature engineering using claims and prescription data, model selection tradeoffs (logistic regression vs gradient boosting given explainability requirements for healthcare), how you'd handle class imbalance (around 10% readmission rate), and how you'd evaluate fairness across demographic groups before deployment.
Ho sostenuto un colloquio presso CVS Health (New York, NY)
Colloquio
Easy Chill, comfortable, wasnt difficult, more on health and pharmacy stuff,
Interview was friednly and accommodating, overall i would say good experince, the process was smooth, But applied with out reffereal
Ho presentato la mia candidatura online. Ho sostenuto un colloquio presso CVS Health nel mese di mar 2026
Colloquio
Interestingly, the first round was with AVP, and he discussed many mathematical concepts.. The second and third rounds were on the same day, for coding and case study. One week later was VP interview with behaviors.