Okay, here's my interview experience with Amazon for an internship position. Unfortunately, I was rejected after the technical round, which I found to be of medium difficulty. Let me walk you through it. # The Process The interview was technical and started with brief introductions. We then jumped straight into two Data Structures and Algorithms (DSA) questions. The first was based on binary search, specifically searching on answers. I managed to solve it partially but struggled with the edge cases. The second question was about finding the Lowest Common Ancestor (LCA) in a binary tree. I solved it, but my solution was a bit complex. # Topics Covered & Difficulty The interview focused on Data Structures and Algorithms. One question involved a medium-difficulty binary search problem where I had to efficiently search for an optimal solution within a range, but I struggled with some edge cases. The second question was a standard tree problem, specifically finding the Lowest Common Ancestor in a Binary Tree. While I solved it, I made it a bit more complex than necessary. You can find the exact wording of the questions I was asked in the dedicated section below. # My Thoughts on the Culture Amazon is one of the best companies in the world, and it definitely has a great company culture. The environment seemed very focused and driven, with a strong emphasis on problem-solving and innovation. # Advice for Others My main advice would be to write clean, well-documented code and explain your thought process clearly. The interviewer specifically told me to improve my code writing and explanation skills. Even if you solve the problem, poor code quality or a lack of clear communication can be a reason for rejection. Focus on making your code easy to understand and articulate your reasoning behind each step. Good luck!

## Interview Overview I attended the interview for the AI Internship Programme at the City Union Bank Centre of Excellence, SASTRA University. The interview was around 15 minutes long and focused mainly on my technical background, machine learning project, SQL knowledge, and basic banking concepts. The interview was conversational and the interviewer asked questions based on the skills and projects mentioned in my profile. This made it important to understand my projects properly rather than simply memorizing theoretical answers. ## My Project Discussion A major part of the interview was based on my Telecom Customer Churn Prediction project. I was asked to explain the project, the problem I was trying to solve, the dataset, preprocessing steps, and the machine learning models I used. I explained how I performed exploratory data analysis, identified missing values, handled categorical variables through encoding, prepared the data for modelling, and divided it into training and testing datasets. I had experimented with multiple machine learning algorithms, including: - Logistic Regression - Decision Tree - Random Forest I also discussed the evaluation metrics used to compare the models, including accuracy, precision, recall, and F1-score. One important takeaway from the project discussion was that you should be able to explain not only what model you used, but also WHY you used it and how you decided which model performed better. ## SQL and Technical Questions The interview also covered SQL and basic technical concepts. Revising SQL before the interview was useful, especially concepts such as filtering data, aggregation, grouping, joins, and writing queries to retrieve meaningful information from tables. The interview also tested my understanding of basic machine learning concepts rather than only asking me to write code. For students preparing for similar interviews, I would recommend being comfortable with: - Python fundamentals - Pandas and NumPy - SQL queries - Machine learning fundamentals - Data preprocessing - Classification algorithms - Model evaluation metrics - Basic statistics - Exploratory Data Analysis ## Banking Concepts Since the internship was related to banking and customer analytics, I was also expected to have an understanding of basic banking concepts. The preparation included concepts related to customer analysis, banking products, transactions, and how data analytics can be used to understand customer behaviour. Having a basic understanding of banking terminology is useful when applying for AI or Data Science roles in the BFSI domain. Technical knowledge alone may not be enough when the internship involves solving business problems using financial or customer data. ## What I Learned One of the biggest lessons from the interview was the importance of understanding your own resume. If you mention a machine learning project, be prepared to explain the complete workflow: 1. What problem are you solving? 2. What dataset did you use? 3. What preprocessing did you perform? 4. How did you handle missing values? 5. How did you encode categorical variables? 6. Which models did you try? 7. Why did you choose those models? 8. Which evaluation metrics did you use? 9. Which model performed best and why? 10. What could you improve in the future? The same applies to programming languages and tools mentioned in your resume. Interviewers can ask questions from any skill you list. ## Preparation Tips For students preparing for AI/ML internships, I would recommend focusing on fundamentals instead of trying to learn too many advanced topics at the last minute. Revise Python, SQL, machine learning algorithms, EDA, preprocessing, and evaluation metrics. At the same time, understand your projects deeply and practice explaining them in simple language. For banking-related AI roles, spend some time learning basic banking concepts and customer analytics terminology as well. Most importantly, don't just memorize definitions. Try to understand how a concept would actually be used to solve a real-world problem. ## Final Takeaway The interview was a good opportunity to connect my academic knowledge and machine learning project experience with a real-world banking use case. It also showed me that for internship interviews, strong fundamentals and the ability to clearly explain your own work are extremely important. If you are preparing for a similar AI/Data Science internship, focus on your projects, Python, SQL, ML fundamentals, and domain-specific concepts. Be honest about what you know, and make sure you can confidently explain everything mentioned on your resume.

# KalkiTech Graduate Engineer Recruitment Experience – 2026 I recently attended the KalkiTech recruitment assessment as part of the campus hiring process. The assessment consisted of 27 multiple-choice questions and 2 coding questions. ## Assessment Pattern The first section consisted of 27 MCQs covering quantitative aptitude, logical reasoning, probability, percentages, profit and loss, compound interest, data interpretation, calendars, puzzles, and logical arrangement problems. Some of the questions I remember included: - A word arrangement question involving grouping vowels together and finding the maximum number of possible arrangements. - A sales tax and percentage-based problem. - A probability question involving red, yellow, and blue balls selected randomly from a bag. - A profit, cost price, marked price, and discount problem. - A percentage problem involving P and Q. - A compound interest problem with quarterly compounding. - A bar graph/data interpretation question. - A number puzzle involving three circles divided into halves. - A logical friendship/relationship problem. - Calendar and day-of-the-week questions. - A problem involving people, professions, relationships, and marriage conditions. There was also a programming-related conceptual question about JavaScript modules and exports. The question tested the difference between named and default exports and how functions can be exposed while keeping internal helper functions private. ## Coding Questions There were 2 coding questions. ### Coding Question 1 – Array / Target Reduction Given a sequence of numbers and a target value, elements can be removed only from the beginning or the end of the sequence. Each removed element decreases the target by its value. The task was to find the minimum number of removal operations required to make the target exactly zero. If it is impossible, the result should be -1. This problem can be approached using prefix sums, sliding window, or related array techniques. ### Coding Question 2 – Grid Minimum Maximum Strain A rectangular grid contains numerical values representing stability ratings. Movement is allowed up, down, left, or right. The cost of moving between two adjacent cells is the absolute difference between their values. Starting from the top-left cell, the goal is to reach any cell in the bottom row while minimizing the maximum movement cost encountered along the route. After finding the minimum possible maximum strain, a fixed value of 2 must be added to the result. This is essentially a minimum-bottleneck path problem on a grid and can be solved using approaches such as Dijkstra's algorithm with a minimax distance, binary search with graph traversal, or related graph techniques. I would rate the assessment as Medium. The aptitude section required good time management because it included a mixture of quantitative, logical, and puzzle-based questions. The coding questions required understanding the underlying problem rather than simply applying basic syntax. For candidates preparing for a similar Kalkitech assessment, I would recommend focusing on: - Percentages, profit and loss, simple and compound interest - Probability - Data interpretation - Calendar and logical reasoning problems - Puzzles and arrangement questions - Arrays and prefix sums - Sliding window - Hashing - Graph and grid problems - Basic SQL and DBMS - OOP concepts - Operating Systems - Programming fundamentals This experience is based on the questions I personally remember from the assessment. The wording of the questions above has been paraphrased rather than reproduced from an assessment paper.