Implement HashMap with Separate Chaining

I would start by defining a Node class containing the key, value, and a reference to the next node. The HashMap will use an array of these nodes as buckets. For the put operation, I calculate the index using the hash code modulo the array length. If the bucket is empty, I create a new node; otherwise, I traverse the linked list. If the key exists, I update the value; if not, I append a new node at the end. The get method follows a similar traversal pattern, returning null if the key isn't found. A critical part of this implementation is the resize logic. When the load factor exceeds a threshold, typically 0.75, I allocate a new array with double the capacity. I then iterate through every existing bucket and re-insert all entries into the new array using the updated modulo calculation. This ensures that even with many collisions, the average time complexity remains O(1). At Amazon, where scalability is paramount, handling the resize efficiently prevents performance degradation during high-volume data ingestion. Finally, I would ensure edge cases like negative hash codes or null keys are handled gracefully to prevent runtime errors in production environments.