Walk me through your production database architecture experience and database selection criteria.

In Simple Words: "In production, I have architected and operated PostgreSQL, MySQL, MongoDB, Redis, and DynamoDB. My database selection framework is guided by access patterns and consistency requirements:

• PostgreSQL: My default standard for transactional services, relational integrity, ACID compliance, and JSONB flexibility.
• MongoDB: For polymorphic documents and aggregates where entities are frequently read and written as complete nested units.
• Redis: For ultra-low-latency in-memory caching, distributed locks, session stores, and rate limiters.
• DynamoDB: For serverless architectures requiring predictable single-digit millisecond latency at arbitrary scale."

-- Production PostgreSQL table with UUIDs, indexing, and JSONB document storage:
CREATE TABLE users (
  id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
  email VARCHAR(255) UNIQUE NOT NULL,
  metadata JSONB NOT NULL DEFAULT '{}',
  created_at TIMESTAMPTZ NOT NULL DEFAULT NOW()
);

CREATE INDEX idx_users_metadata_gin ON users USING GIN (metadata);