Hello. I have a multi TB relational data in PostgreSQL. I am in search for better analytics query performance for our clients. We have some constraints while achieving this. • We need fresh data on the analytics database. So we can't tolerate delays more than ~5 seconds on the analytics site. • We may want to denormalize some tables on the analytics side. When the storage is columnar I am not sure how denormalization would help? • PostgreSQL is hosted on AWS. So if possible we want to host the analytical engine be hosted on AWS too. What options do we have for running Druid on the cloud?

Hi Yusuf, Denormalization - this works for any relational database product ... in the case of Druid it can eliminate joins and/or summarize data. Druid doesn't necessarily need that ... but depending on the use case it could significantly improve query performance. AWS - Druid runs will on AWS instances, both Intel and Graviton based. However if you are looking for a fully hosted SaaS Druid offering you may want to check out Imply Polaris. Thanks. John