What do you like best about the product?

There are so many things but for starters, it's just how fast we are able to serve up query results to applications and end users regardless of the data set size. Hugely complicated queries run like lightning fast, so much so that we've pivoted from using Firebolt just for reporting and analytics to also including it in our internal tooling for our business. Things we simply couldn't pull off otherwise without a massive scaling of our infrastructure.

The cost is simply unmatched for the levels of feature and performance. We're a small company with big ambitions but cost is still a paramount consideration. We have flexibility in controlling our costs and scaling as we see fit, which is great.

I've found that my team has generally found the syntax to be pretty easy to pick up, it shares a lot of commonality with other SQL flavors we're familiar with. And even when it stumps us, the documentation is pretty robust (unlike many software vendors we use) and their support staff has been top-notch. They even share a channel in our Slack where we get quick answers to many of our questions. Firebolt goes so far as to proactively watch our query performance patterns and shares that with me and will help us fix them.

Simply put, I work WITH Firebolt to solve our problems and limitations instead of working AROUND them like we have to with some other vendors.

What do you dislike about the product?

The segregation of ingestion (or "general purpose") engines from read-only "analytics" engines is great from a granular control mechanism, especially as it pertains to cost. However that often requires bouncing of the read-only engine that services our applications to reflect changes made by the ingestion engine, causing a several minute outage. There are ways around it but this is often the case for us at this time.

Additionally, it would be great if there was a way to have stored queries accessible via the API access. Currently the query has to be passed to the API meaning that even where the same query could be used in our code, that is being duplicated and would necessitate an application deployment to make alterations. Compare that to a stored function/procedure model in our sql transaction engine and this is not ideal for a small shop that moves this quickly.

What problems is the product solving and how is that benefiting you?

Firebolt serves us as our data warehouse, our primary ETL processing engine, reporting and analytics platform as well as our high performance query engine for internal tools built for the rest of our organization.