Backend Developer Interview Questions (Practice with AI Feedback)

"For high-write scenarios, I minimize indices on write-heavy tables since indexing slows writes. I separate database reads and writes using a primary-replica replica model where writes go to the primary node and reads are distributed to replicas. I implement Redis caching for hot read paths, and shard the database horizontally based on a tenant ID if the scale exceeds a single instance."

"SQL databases are relational, use structured schemas, and guarantee strict ACID transactions, making them ideal for financial systems or complex join operations. NoSQL databases are non-relational, offer dynamic schemas, and scale horizontally by design. I choose NoSQL (like MongoDB or DynamoDB) when handling unstructured data, high-velocity streams, or when horizontal scaling is a priority."

"I secure distributed APIs using OAuth 2.0 and JWT. When a user logs in, the authentication service issues a signed cryptographical JWT. The client sends this token in the Authorization header. An API Gateway validates the signature using public keys (JWKS) before routing requests. For microservice-to-microservice calls, I use mutual TLS (mTLS) to encrypt and verify the identity of each service."