Hi @Arindam Halder​ , Delta lake is more performant compared to a regular parquet table. pls check below for some stats on the performance

https://docs.azuredatabricks.net/_static/notebooks/delta/optimize-python.html

yes, you can use it for time series forecasting. You could watch the below video to see how forecasting is done at scale using Delta lake

https://databricks.com/session_na20/real-time-forecasting-at-scale-using-delta-lake-and-delta-cachin...

you could also find the delta streaming APIs here

https://docs.databricks.com/delta/delta-streaming.html#table-streaming-reads-and-writes