How would you design a time series database for product sales and store data, and how would you retrieve daily, monthly, and weekly summaries?

A time series database is optimized for storing and querying data points indexed by time. For product sales and store data, you would model each record with a timestamp, store_id, product_id, and sales_amount. Using a database like TimescaleDB or Amazon Timestream, you can create a hypertable or time‑partitioned table. Index the timestamp and any tags (store_id, product_id) for fast filtering. For daily summaries, use a query such as SELECT time_bucket('1 day', ts) AS day, SUM(sales_amount) FROM sales GROUP BY day; For monthly summaries, replace '1 day' with '1 month'; for weekly, use '1 week'. You can also create continuous aggregates to materialize these summaries. Compression and retention policies help manage storage. This design allows efficient ingestion, real‑time analytics, and scalable querying across large volumes of sales data. Additionally, you can use downsampling and roll‑up tables to keep query latency low while retaining historical detail.