Tiger Analytics Data Scientist interview typically runs 3-4 rounds: technical, deeper technical/project discussion, and HR, sometimes with an extra project or salary round. It usually takes a few weeks and is notably structured and fundamentals-focused.
$123K
Avg. Base Comp
$155K
Avg. Total Comp
3-4
Typical Rounds
2-4 weeks
Process Length
Our candidates consistently describe Tiger Analytics as a process that looks simple on the surface but gets much sharper once the conversation turns to real work. The recurring theme is fundamentals plus proof: interviewers want to see that you can handle core Python, SQL, and ML concepts cleanly, but they also keep pulling the discussion back to what you actually built. Multiple candidates noted that the most important conversations were not the ones with the flashiest questions, but the ones where they had to walk through preprocessing, feature engineering, model selection, and the tradeoffs behind those choices.
What makes this company different is how much weight they place on business context and implementation detail. Our candidates report that even when the technical questions were straightforward, the follow-up was often scenario-based and CV-driven, with interviewers probing whether the work had real impact and whether the candidate could defend every line on the resume. We also see a pattern of practical coding rather than algorithmic trickery: simple list, dictionary, pandas, and SQL tasks show up alongside basic ML theory. That combination tells us Tiger Analytics is screening for people who can move comfortably between analysis, coding, and client-facing explanation without hiding behind buzzwords.
Synthetized from 3 candidates reports by our editorial team.
Had an interview recently?
Share your experience. Unlock the full guide.
Real interview reports from people who went through the Tiger Analytics process.
The interview focused heavily on one of my projects. I was first asked to explain the project, and then the interviewer went deeper into the end-to-end deployment work.
They asked what steps I followed during EDA, what kind of analysis I did, why I chose the model I used, and what techniques I used for missing values. I mentioned imputation and MICE. They also asked how I would handle an imbalanced dataset. After that, they asked me to explain the differences between Random Forest, XGBoost, and LightGBM. There were also questions on classification metrics, especially the AUC-ROC curve and how I would explain it clearly.
I felt fairly confident explaining my project flow and the model-related questions because I had worked on those properly. I struggled more when they went deeper into deployment and asked how exactly the model was deployed, what tools I used, and how I checked for data drift and model drift after deployment.
They also asked about the bias-variance trade-off. Apart from ML questions, there was one Python coding question where I had to check whether one string, like "aab", is present inside another string, like "bacaabacd". There was also an SQL question to find employees whose salary is greater than the average salary.
Overall, it was not just a basic project discussion. They picked one project and went end-to-end, from EDA to model selection, evaluation, deployment, and monitoring. Be ready to explain your project properly, not just the surface-level points.
Share your own interview experience to unlock all reports, or subscribe for full access.
Sourced from candidate reports and verified by our team.
Topics based on recent interview experiences.
Featured question at Tiger Analytics
Select the 2nd highest salary in the engineering department
| Question | |
|---|---|
| Closest SAT Scores | |
| Experiment Validity | |
| Prime to N | |
| Find the Missing Number | |
| Bagging vs Boosting | |
| Get Top N Frequent Words | |
| New Partner Card | |
| Minimum Absolute Distance | |
| Missing Housing Data | |
| Target Indices | |
| Median O(1) | |
| Assumptions of Linear Regression | |
| Digit Accumulator | |
| Bias - Variance Tradeoff and Class Imbalance in Finance | |
| Matrix Rotation | |
| Count Transactions | |
| Transformer Encoder Layer | |
| KNN From Scratch | |
| Possible Triangles | |
| Yelp-like System | |
| Production Model Monitoring | |
| Data Preparation for Imbalanced Data | |
| Finding the Maximum Number in a List | |
| String Palindromes | |
| Minimum Directional Path | |
| Normal Distribution Sample | |
| k-Means from Scratch | |
| Area Under the ROC Curve | |
| Maximum Common Substring |
Synthesized from candidate reports. Individual experiences may vary.
The first interview is a fundamentals-heavy technical screen focused on core Python, SQL, and basic machine learning concepts. Candidates are commonly asked about data handling, loops, dictionaries, joins, aggregations, window functions, and ML basics like supervised vs. unsupervised learning, bias-variance, and evaluation metrics.
This round goes deeper into your past work and technical decision-making. Interviewers often dig into end-to-end projects, including preprocessing, feature engineering, model selection, evaluation, and the challenges you faced, along with more advanced topics like NLP, transformers, and practical SQL or Python questions.
Some candidates reported a separate round dedicated almost entirely to resume and project discussion. This stage is highly CV-driven and focuses on the business impact of your work, so you should be ready to explain every line on your resume and the tradeoffs behind your technical choices.
The final stage is an HR conversation covering behavioral fit, communication, career goals, and overall alignment with the team. In some cases, this round also includes salary discussion and is typically much lighter than the technical interviews.