Absolutdata Applied Scientist Interview: Rounds, Questions and Preparation

What the job involves

Applied Scientists in AI and deeptech build and deploy machine learning models to solve business problems, often working on recommendation systems, forecasting, or natural language processing. At fast-growing firms like Absolutdata, they collaborate with data engineers and product teams to turn prototypes into production-ready solutions. The role blends research, coding, and client-facing communication, requiring both theoretical depth and practical implementation skills.

Role and Technical Knowledge

1. 1

   Explain the difference between bagging and boosting, and give an example of when you would use each.

   Cover variance reduction vs. bias reduction, mention Random Forest for bagging and XGBoost for boosting, and relate to a business problem like churn prediction.

2. 2

   How would you handle missing data in a dataset for a predictive model?

   Discuss deletion, imputation (mean, median, mode), and advanced methods like KNN imputation, and mention when each is appropriate based on missingness pattern.

3. 3

   What is the bias-variance tradeoff, and how do you detect and address it?

   Explain underfitting vs. overfitting, use learning curves, and suggest regularization, more data, or simpler models as remedies.

4. 4

   Describe a time you used a statistical test to validate a hypothesis in a project.

   Use a concrete example like A/B testing with t-test or chi-square, state the null hypothesis, p-value, and business decision made.

Practical Scenarios

1. 5

   You are given a dataset of customer transactions and asked to build a recommendation system. Outline your approach.

   Start with data exploration, choose collaborative filtering or content-based methods, discuss evaluation metrics like precision@k, and mention cold-start handling.

2. 6

   A model performs well in testing but poorly in production. What could be the reasons and how would you fix it?

   Consider data drift, leakage, overfitting, or pipeline bugs, and suggest monitoring, retraining, and validation on recent data.

3. 7

   How would you deploy a machine learning model as an API for a real-time application?

   Mention Flask/FastAPI, containerization with Docker, cloud services like AWS SageMaker, and load balancing for scalability.

4. 8

   You need to reduce model inference time without sacrificing much accuracy. What techniques would you use?

   Discuss model pruning, quantization, using lighter models (e.g., MobileNet), and caching or batch processing.

Behavioural and HR

1. 9

   Tell me about a time you worked on a team project where you had to explain a complex technical concept to a non-technical stakeholder.

   Use STAR method, highlight simplification, visual aids, and the positive outcome like stakeholder approval or project success.

2. 10

   Why do you want to work as an Applied Scientist at a fast-growing AI company like Absolutdata?

   Show passion for applied AI, mention the company's focus areas, and align with your career goals of solving real-world problems.

3. 11

   Describe a situation where you had to learn a new technology or tool quickly to complete a task.

   Give a specific example, mention resources used, time frame, and how it contributed to project success.

4. 12

   How do you prioritize tasks when working on multiple projects with tight deadlines?

   Discuss using urgency-importance matrix, communication with managers, and breaking tasks into smaller milestones.

Coding and Problem Solving

1. 13

   Write a Python function to compute the F1 score given true and predicted labels.

   Show understanding of precision, recall, and harmonic mean, and handle edge cases like zero division.

2. 14

   Given a list of numbers, find the pair that sums to a target value. Optimize for time complexity.

   Use a hash map for O(n) solution, explain trade-offs with brute force, and mention space complexity.

3. 15

   How would you implement a simple linear regression from scratch using gradient descent?

   Cover hypothesis function, cost function, gradient computation, and update rule, and mention learning rate tuning.

4. 16

   Write an SQL query to find the second highest salary from an employee table.

   Use subquery with LIMIT/OFFSET or window function like DENSE_RANK, and explain handling of ties.