Cracking the Tiger Analytics Interview
Tiger Analytics is one of the fastest-growing advanced analytics consulting firms in India (headquartered in Chennai/US). They look for hands-on, highly technical data scientists who can write production-ready code.
The Interview Stages
Tiger Analytics is known for having a very structured and rigorous technical screening process.
Round 1: Hackerrank / Coderpad Test
- Focus: Heavy SQL and Python (Pandas/NumPy) manipulation.
- Task: You might be given a messy dataset and asked to clean it, handle missing values, and aggregate the data to find specific business insights within 60 minutes.
Round 2: Technical Interview (Statistics & ML)
- Statistics: "Explain Central Limit Theorem." "How do you check if a variable is normally distributed?"
- ML Algorithms: "How do you decide the number of clusters 'K' in K-Means clustering?" (Discuss the Elbow method or Silhouette score).
- Imbalanced Data: "How do you handle a dataset with 99% negative cases and 1% positive cases?" (Discuss SMOTE, adjusting class weights, and using precision/recall instead of accuracy).
Round 3: Business Case & ML Deployment
- Scenario: "Build a churn prediction model for a telecom company."
- The Differentiator: The interviewer will ask, "Once the model is built in your Jupyter Notebook, how do we get the sales team to actually use it?"
- The Answer: Discuss deploying the model as a REST API using Flask/FastAPI, containerizing it with Docker, and pushing the predictions into a Tableau dashboard for the sales team.
Round 4: HR / Director Round
- Tiger values a strong engineering culture. They will check your cultural fit and your willingness to learn new cloud tools (AWS/Azure) on the fly for specific client projects.
Final Advice: Do not fake projects on your resume. Tiger interviewers will drill down to the absolute microscopic details of any project you list, asking about your specific hyperparameter tuning decisions and why you chose a specific loss function.
