Ho presentato la mia candidatura online. La procedura ha richiesto 3 settimane. Ho sostenuto un colloquio presso Boston Consulting Group (Casablanca, ) nel mese di ott 2025
Colloquio
Le processus a commencé par un test technique sur CodeSignal avec des exercices de programmation et d'analyse de données assez complexes. Ensuite, j'ai eu un entretien avec les RH pour discuter de mon parcours et de mes motivations. Le niveau de difficulté du test CodeSignal était élevé, avec des questions avancées en SQL, Python et résolution de problèmes algorithmiques. Durée totale du processus : environ 2-3 semaines.
Domande di colloquio [1]
Domanda 1
Test CodeSignal avec 3 notebooks : analyser et nettoyer des datasets avec Pandas/NumPy (gestion des valeurs manquantes, groupby, merge), puis implémenter et évaluer un modèle d'arbre de décision avec des métriques de performance
Ho sostenuto un colloquio presso Boston Consulting Group (Bengaluru)
Colloquio
I attended the initial online assessment, where there were MCQ questions on statistics, probability, machine learning, and three coding problems on Python. We had to do data analysis, EDA, feature engineering, and fitting the model and creating a prediction file.
Domande di colloquio [1]
Domanda 1
Simple ML questions:
- Bagging
- Boosting
- Logistic Regression
- Random Forest
- Decision Trees
Ho presentato la mia candidatura online. Ho sostenuto un colloquio presso Boston Consulting Group (Bogotá, Bogota) nel mese di mar 2026
Colloquio
The interviewer was very friendly, he explained everything I think he did not leave anything out in terms of salary and benefits, he also explained what would be needed for the technical assessment. In my case, Python (Numpy, Pandas, Scikit-learn).
Ho presentato la mia candidatura online. Ho sostenuto un colloquio presso Boston Consulting Group (Casablanca, ) nel mese di feb 2026
Colloquio
I applied online through the BCG careers page. After about a week, I received an email inviting me to complete a proctored CodeSignal Data Science Framework assessment. The assessment was 90 minutes long and contained 5 questions covering data cleaning, data preprocessing, training a ML classifier, aggregating data from multiple files using joins and groupbys, and a probability calculation. The time pressure was significant. I did not advance past this stage.
Domande di colloquio [1]
Domanda 1
Given a dataset of drivers with multiple CSV files, clean the data, preprocess features using scikit-learn (imputation, encoding, scaling), and train a classifier to predict a driver's class.
