Accenture
Global professional services company with leading capabilities in digital, cloud and security.
4 Rounds
~21 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
Data Scientist
•
Behavioral
•
medium
Tell me about a time you had to explain a complex machine learning model to a non-technical client stakeholder.
#Stakeholder Management
#Storytelling
#Consulting
Data Scientist
•
Behavioral
•
hard
Describe a situation where a client's data maturity was much lower than expected. How did you adjust the data science project scope?
#Scope Management
#Client Expectations
#Data Engineering
Data Scientist
•
Behavioral
•
medium
Tell me about a time you disagreed with a senior technical lead or a client regarding the choice of an algorithm. How did you resolve it?
#Conflict Resolution
#Influence
#Evidence-Based Decisions
Data Scientist
•
Behavioral
•
medium
Accenture frequently requires consultants to switch tech stacks based on client needs. Tell me about a time you had to rapidly learn a new technology to deliver a project.
#Continuous Learning
#Consulting
#Agility
Data Scientist
•
Behavioral
•
hard
What would you do if you discovered that the model you built for a client exhibits significant bias against a protected demographic?
#Responsible AI
#Ethics
#Client Communication
Data Scientist
•
Behavioral
•
medium
Tell me about a time when a client kept adding features to a machine learning project (scope creep). How did you handle it?
#Scope Creep
#Agile
#Client Management
Data Scientist
•
Behavioral
•
easy
Why are you interested in joining Accenture as a Data Scientist, and how does consulting differ from working in a traditional product company?
#Motivation
#Consulting Mindset
#Career Goals
Data Scientist
•
Coding
•
medium
Write a SQL query to calculate the 7-day rolling average of daily transaction volumes for a specific client account.
#Window Functions
#Time Series
#Data Aggregation
Data Scientist
•
Coding
•
easy
Given a Pandas DataFrame with a million rows of client sales data, how would you efficiently find the top 5% of customers by revenue, handling missing values?
#Python
#Pandas
#Data Cleaning
Data Scientist
•
Coding
•
medium
Write a Python function to implement K-Means clustering from scratch. You can use NumPy but not scikit-learn.
#Python
#Machine Learning Algorithms
#NumPy
Data Scientist
•
Coding
•
medium
Write a SQL query using window functions to find the top 3 highest-paid employees in each department of a client's organization.
#Window Functions
#DENSE_RANK
#Joins
Data Scientist
•
Coding
•
easy
Given an array of integers, write a function to return the indices of the two numbers that add up to a specific target. (Two Sum)
#Hash Maps
#Arrays
#Optimization
Data Scientist
•
Coding
•
medium
Write a Python script to compute the TF-IDF matrix for a list of text documents without using external NLP libraries.
#NLP
#Python
#Math
Data Scientist
•
Coding
•
medium
Write a SQL query to identify customers who have made a purchase in every single month of the year 2023.
#Aggregation
#Date Functions
#HAVING Clause
Data Scientist
•
Coding
•
medium
Write a function to find the longest substring without repeating characters in a given string.
#Sliding Window
#Strings
#Hash Maps
Data Scientist
•
Coding
•
hard
Write a Python function to merge k sorted arrays. What is the time and space complexity of your solution?
#Heaps
#Priority Queue
#Sorting
Data Scientist
•
System Design
•
hard
Design an end-to-end architecture for a real-time credit card fraud detection system on AWS or Azure.
#Streaming Data
#Cloud Architecture
#Latency
#MLOps
Data Scientist
•
System Design
•
medium
A retail client wants to implement a personalized product recommendation engine. Walk me through the high-level system design.
#Recommendation Systems
#Collaborative Filtering
#Scalability
Data Scientist
•
System Design
•
hard
Design a system architecture for an internal HR chatbot that securely queries proprietary company policies using an LLM.
#LLMs
#Security
#Vector Databases
#RAG
Data Scientist
•
System Design
•
medium
How would you design a pipeline to detect data drift and model decay for a pricing model deployed in production?
#MLOps
#Model Monitoring
#Data Drift
Data Scientist
•
System Design
•
hard
Design a scalable data ingestion and preprocessing pipeline for a client receiving 500GB of unstructured text logs daily.
#Data Engineering
#Big Data
#Spark
#Cloud Storage
Data Scientist
•
System Design
•
medium
A client wants to test a new machine learning-based search ranking algorithm. Design the A/B testing framework and define the success metrics.
#A/B Testing
#Experimentation
#Metrics
Data Scientist
•
System Design
•
hard
Design an automated system for a healthcare client to extract and structure data from scanned medical invoices using OCR and NLP.
#Computer Vision
#OCR
#NLP
#Information Extraction
Data Scientist
•
System Design
•
medium
Walk me through the steps to containerize a trained scikit-learn model using Docker and deploy it as a REST API on AWS ECS or Azure Container Instances.
#Docker
#Cloud Deployment
#APIs
#FastAPI/Flask
Data Scientist
•
Technical
•
medium
A client wants to predict customer churn but their dataset is highly imbalanced (99% retain, 1% churn). How do you approach this?
#Imbalanced Data
#Classification
#SMOTE
#Evaluation Metrics
Data Scientist
•
Technical
•
hard
For a recent enterprise knowledge base project, how would you decide between fine-tuning an open-source LLM versus using a Retrieval-Augmented Generation (RAG) approach?
#LLMs
#RAG
#Fine-tuning
#NLP
Data Scientist
•
Technical
•
medium
Explain the mathematical difference between L1 and L2 regularization. When would you recommend one over the other to a client?
#Regularization
#Lasso
#Ridge
#Feature Selection
Data Scientist
•
Technical
•
hard
Explain how self-attention mechanisms work in Transformer models. Why are they more efficient than RNNs for long-context client documents?
#Transformers
#NLP
#Attention Mechanism
Data Scientist
•
Technical
•
medium
Your classification model has a 95% accuracy, but the client is unhappy because it's missing critical fraudulent transactions. What metric should you have optimized instead?
#Evaluation Metrics
#Recall
#Fraud Detection
Data Scientist
•
Technical
•
medium
Explain how Gradient Boosting works. How does it differ from Random Forest?
#Ensemble Methods
#XGBoost
#Random Forest
Data Scientist
•
Technical
•
hard
A supply chain client needs you to forecast inventory demand. Compare the pros and cons of using ARIMA versus an LSTM network for this task.
#Forecasting
#ARIMA
#Deep Learning
#LSTM
Data Scientist
•
Technical
•
medium
Explain the assumptions of linear regression. What happens if the assumption of homoscedasticity is violated?
#Linear Regression
#Statistical Assumptions
#Econometrics
Data Scientist
•
Technical
•
easy
What is the curse of dimensionality, and what techniques would you use to overcome it in a high-dimensional client dataset?
#Dimensionality Reduction
#PCA
#Feature Engineering
Data Scientist
•
Technical
•
medium
How do you handle categorical variables with high cardinality (e.g., zip codes) in a tree-based model versus a neural network?
#Categorical Encoding
#Embeddings
#Target Encoding
Data Scientist
•
Technical
•
hard
You have deployed a deep learning model for a client, but the inference latency is too high. What techniques would you use to optimize the model's inference speed?
#Model Quantization
#Pruning
#ONNX
#TensorRT
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.