Infosys
Global leader in next-generation digital services and consulting.
3 Rounds
~14 Days
Medium
The Interview Loop
Recruiter Screen (30 min)
Standard fit check, behavioral questions, and resume overview.
Technical Loop (3-4 Rounds)
Deep dive into domain knowledge, coding, and system design.
Interview Question Bank
Machine Learning Engineer
•
Behavioral
•
medium
Tell me about a time you had to explain a complex machine learning model to a non-technical client stakeholder. How did you ensure they understood?
#Client Interaction
#Soft Skills
Machine Learning Engineer
•
Behavioral
•
hard
Describe a situation where your machine learning model performed exceptionally well in training but failed or underperformed in production. How did you troubleshoot and fix it?
#Debugging
#Experience
Machine Learning Engineer
•
Behavioral
•
easy
Working at Infosys often means handling multiple client deliverables with tight deadlines. How do you prioritize your ML development tasks?
#Agile
#Prioritization
Machine Learning Engineer
•
Behavioral
•
medium
Tell me about a time you disagreed with a senior engineer or a client regarding the choice of an ML algorithm or architecture. How did you resolve it?
#Teamwork
#Leadership
Machine Learning Engineer
•
Behavioral
•
easy
The AI landscape, especially GenAI, is changing rapidly. How do you stay updated with the latest tools, frameworks, and research papers?
#Self-Improvement
#AI Trends
Machine Learning Engineer
•
Behavioral
•
medium
Describe a time when you had to work with messy, unstructured, or incomplete data provided by a client. How did you approach the data cleaning and structuring process?
#Data Engineering
#Client Interaction
Machine Learning Engineer
•
Coding
•
easy
Given an array of integers and an integer target, return indices of the two numbers such that they add up to target. Can you do it in O(n) time?
#Arrays
#Hash Map
Machine Learning Engineer
•
Coding
•
medium
Write a SQL query to find the nth highest salary from an Employee table. How would you handle ties?
#SQL
#Window Functions
Machine Learning Engineer
•
Coding
•
medium
Write a Python function to calculate the TF-IDF scores for a given corpus of text documents from scratch, without using scikit-learn.
#Python
#NLP
#Math
Machine Learning Engineer
•
Coding
•
medium
Given a list of intervals representing client server downtimes, merge all overlapping intervals.
#Arrays
#Sorting
Machine Learning Engineer
•
Coding
•
medium
Write a SQL query using window functions to calculate the 7-day rolling average of transaction volumes for a banking client.
#SQL
#Time Series
#Window Functions
Machine Learning Engineer
•
Coding
•
hard
Implement the K-Means clustering algorithm from scratch in Python using NumPy.
#Python
#NumPy
#Algorithms
Machine Learning Engineer
•
Coding
•
medium
Find the length of the longest substring without repeating characters in a given string.
#Strings
#Sliding Window
Machine Learning Engineer
•
Coding
•
easy
Write a Python script using Pandas to handle missing values in a dataset containing both numerical and categorical columns, applying mean imputation for numerical and mode for categorical.
#Python
#Pandas
#Data Cleaning
Machine Learning Engineer
•
System Design
•
hard
Design a predictive maintenance system for a manufacturing client's IoT sensor data to predict machine failures before they happen.
#IoT
#Time Series
#Architecture
Machine Learning Engineer
•
System Design
•
hard
How would you design a real-time credit card fraud detection system capable of processing thousands of transactions per second?
#Real-time Processing
#Scalability
#Machine Learning
Machine Learning Engineer
•
System Design
•
hard
Design a scalable recommendation engine for an e-commerce client. How do you handle the cold start problem for new users and new items?
#Recommender Systems
#Architecture
Machine Learning Engineer
•
System Design
•
medium
Design an MLOps pipeline for the continuous training and deployment of an NLP sentiment analysis model.
#MLOps
#CI/CD
#NLP
Machine Learning Engineer
•
System Design
•
medium
A client wants to automate invoice processing. Design a document extraction system using OCR and Large Language Models.
#Computer Vision
#LLMs
#Architecture
Machine Learning Engineer
•
System Design
•
hard
How do you serve a PyTorch model in production to handle 10,000 requests per minute efficiently?
#Model Serving
#Scalability
Machine Learning Engineer
•
System Design
•
hard
Design a personalized, secure chatbot for a banking client using Generative AI. How do you ensure it doesn't hallucinate sensitive financial advice?
#Generative AI
#Security
#Architecture
Machine Learning Engineer
•
Technical
•
medium
Explain the bias-variance tradeoff. How do Random Forest and Gradient Boosting algorithms handle this tradeoff differently?
#Ensemble Methods
#Model Evaluation
Machine Learning Engineer
•
Technical
•
medium
In a recent fraud detection project for a financial client, the dataset had a 99:1 class imbalance. How would you handle this highly imbalanced dataset?
#Imbalanced Data
#Classification
Machine Learning Engineer
•
Technical
•
hard
Explain the Self-Attention mechanism in Transformer models. Why is it more efficient than RNNs for long sequences?
#NLP
#Transformers
#LLMs
Machine Learning Engineer
•
Technical
•
medium
What is the difference between L1 (Lasso) and L2 (Ridge) regularization? When would you choose one over the other in a consulting project?
#Regularization
#Linear Models
Machine Learning Engineer
•
Technical
•
medium
How do you evaluate a binary classification model? Explain the ROC-AUC curve and the Precision-Recall tradeoff.
#Metrics
#Statistics
Machine Learning Engineer
•
Technical
•
hard
Explain how XGBoost works under the hood. What makes it faster and more accurate than traditional Gradient Boosting?
#Ensemble Methods
#XGBoost
Machine Learning Engineer
•
Technical
•
medium
What are the vanishing and exploding gradient problems in deep neural networks, and how do LSTMs or ResNets solve them?
#Neural Networks
#RNN
#CNN
Machine Learning Engineer
•
Technical
•
medium
How does Retrieval-Augmented Generation (RAG) work? Walk me through how you would implement it for an enterprise knowledge base.
#LLMs
#RAG
#Vector Databases
Machine Learning Engineer
•
Technical
•
medium
What is the difference between fine-tuning an LLM (like Llama 3) and using Prompt Engineering with RAG? When would you recommend each to a client?
#LLMs
#Fine-tuning
#Architecture
Machine Learning Engineer
•
Technical
•
medium
What is cross-validation, and why is standard k-fold cross-validation inappropriate for time-series data? What should you use instead?
#Time Series
#Statistics
Machine Learning Engineer
•
Technical
•
medium
Explain the concept of Word2Vec. What is the difference between the Continuous Bag of Words (CBOW) and Skip-gram architectures?
#NLP
#Embeddings
Machine Learning Engineer
•
Technical
•
hard
How do you detect data drift and concept drift in a deployed machine learning model? How do you mitigate it?
#Model Monitoring
#Data Drift
Machine Learning Engineer
•
Technical
•
easy
What are the core assumptions of Linear Regression? How do you check if they are violated?
#Linear Models
#Statistics
Machine Learning Engineer
•
Technical
•
medium
Explain the mathematical intuition behind Principal Component Analysis (PCA). How do you decide the number of principal components to keep?
#Dimensionality Reduction
#Math
Difficulty Radar
Based on recent AI-sourced data.
Meet Your Interviewers
The "Standard" Interviewer
Senior EngineerFocuses on core competencies, system constraints, and clear communication.
SimulateUnwritten Rules
Think Out Loud
Always explain your thought process before writing code or drawing architecture.