Amazon

Amazon Data Scientist Interview Experience Round 1 DSA Questions: You are given an array with daily temperatures and you have to find after how many days you will find a temperature greater than your current temperature. It was a monotonic stack question. ML round: Asked about my thesis. RAG. Its working architecture. And some situation questions such as if the retrieval is not correct, how will you fix it? How will you evaluate it? What is the chunking you did? Bias-variance tradeoff, why it happens. Why we cannot lower it. Regularisation in depth. Dropout in depth. Round 2 Project discussion on my most recent project. It was a RAG project. Then after that they changed the requirement and told me to re-design the architecture of the model. And then some cross questions on why this choice and not something else. Regularization. What we do at the time of inference in dropout. And what if we don't multiply the weights with the probability, and use the same method we used while training. How the prediction will change. Batch normalization. And how does it affect our model. What is the thing that separates the Isolation Forest from the Random Forest. T-test, P-test, A/B testing. Bonus Question: These were asked from my friend Class imbalance. What will you handle it? Tell the end-to-end process. Whether the customer will return the item or not. How can you use the AI/ML knowledge in this? Best Time to Buy and Sell Stock II. Find the top 5 users by total spending in the last six months. For each category find the top-selling product.