When would you use Snowflake, Redshift, and tasks?

Snowflake and Amazon Redshift are both cloud data warehouses, but they differ in architecture and use cases. Snowflake’s multi‑cluster shared data architecture allows independent scaling of compute and storage, making it ideal for workloads with unpredictable or bursty query patterns, and for workloads that require concurrent users without performance contention. Redshift, built on a Massively Parallel Processing (MPP) model, excels in large, steady‑state analytical workloads and offers tight integration with the AWS ecosystem. Use Snowflake when you need elastic scaling, zero‑maintenance, and support for semi‑structured data (JSON, Avro, Parquet). Use Redshift when you have a large existing AWS stack, need cost‑effective scaling for large data volumes, and can tolerate a single cluster’s concurrency limits. "Tasks" in Snowflake refer to scheduled SQL statements that automate ETL/ELT workflows; they are useful for orchestrating data pipelines, refreshing materialized views, or running periodic maintenance. Redshift does not have an equivalent built‑in task scheduler; you would typically use AWS Glue, Step Functions, or external cron jobs to schedule Redshift queries.