CBA

Role: Data Scientist

College: Not Specified (On-Campus Placements)

Interview Date: January 11, 2025

Interview Type: On-Campus Placements

Result: Selected (Inferred)

Difficulty: Medium to Hard

Rounds: 4 rounds (Online Assessment + 3 Interview Rounds)

Topics Asked: Aptitude, Statistics, Machine Learning, Coding (Trees, PCA, Neural Network Debugging), ML Models, Regularization, CNN, Bayes Rule, Case Study, Resume Analysis, Supervised/Unsupervised Learning, K-Means Clustering, HR (internships, co-curriculars, GenAI debate, handling disagreements)

Experience:

This interview experience for a Data Scientist role at Commonwealth Bank of Australia (CBA) involved an online assessment followed by three interview rounds.

Online Assessment: The assessment had three sections.

1. **Aptitude MCQs:** General aptitude-based multiple-choice questions.
2. **MCQs on Statistics and Machine Learning:** Topics included Normal Distribution, Variance Calculation, ML Models, and Neural Networks.
3. **Coding Questions:**
   • **DSA Question:** Solve a problem on trees to calculate the sum of node values with more than one child (Python or R only).
   • **PCA Implementation:** Perform PCA on a dataset using only NumPy operations.
   • **Neural Network Debugging:** Fix logical errors in a given neural network class.

Out of 160 students, 25 were shortlisted for interviews.

Round 1: Technical Interview This round was with a Senior Data Scientist and lasted 45-50 minutes. Topics discussed included ML Models (working, advantages/disadvantages, and suitable datasets), L1 vs L2 Regularization (differences and applications), CNN Structure (explain each step, methods to reduce overfitting, and reasoning for each action), and Bayes Rule (a simple application-based question). A case study was also presented: "Predict customer behavior for a new product launch by: Choosing the best ML model; Explaining EDA, data cleaning, and feature engineering processes." Half of the candidates progressed to the next round.

Round 2: Technical Interview Conducted by a Senior Data Science Manager, this round also lasted 45-50 minutes. It involved a deep dive into technical keywords on the resume and prior internship experiences. Other topics included Supervised vs Unsupervised Learning (differences and use cases) and K-Means Clustering explanation.

Round 3: HR Interview This round included discussions on previous internships and co-curricular activities. A debate on "GenAI is a fad" (requiring opposing views) was part of the round. Questions on handling disagreements with managers in prior roles and other situational/random HR questions were asked.

Key takeaways provided by the candidate include being thorough with ML concepts, statistics, and coding fundamentals, being prepared to defend resume details with real-world examples, approaching interviews with a positive mindset, and focusing on clear communication and genuineness for HR rounds.