Data warehousing is typically a rich guys' toy. Large enterprises are only able to leverage data warehouses for data analytics purposes. We wanted to change that and wanted to build a data warehouse template model for businesses across industries.

If Snowflake was not around, we would have used Google's Big Query or Amazon's Redshift, or a MYSQL/Postgres database in a Windows VM (virtual machine). However, Snowflake made it a lot easier for us with loads of features such as encryption of data in motion and at rest, masking policies, time travel (to correct data load issues), controlled access based on roles, data sharing, third-party data from marketplaces, etc.

It helped us to build MVP (minimum viable product) for our idea of building a data warehouse model for small businesses.

About ten years ago, force.com from salesforce.com offered a similar platform for us to build data warehouses. However, our staff with a data engineering background found it easier to build the data warehouse in Snowflake, with the easy-to-use SQL interface and RBAC models (role-based access control). The platform saved us money as it automatically shuts down the compute engines after about five minutes of idle time. Per second billing (above the first minute) is great.

In my view, cost optimization for the computing power required by the ETL jobs, reports, and dashboards is the most valuable feature. Especially for startups, this helps us to keep cost spending within control without having to worry about manually shutting down the server when not used.

As a Google partner, we like to leverage GCP (Google Cloud Platform). Snowflake supports GCP, AWS & Azure platforms. This works just fine for us. Encryption of data with multiple keys for both data in transit and data at rest gives us enough confidence to use snowflake for our customer 360 solutions.

Currently, we use Snowsight only to monitor the usage of the Snowflake environment by our users. However, if Snowsight can be improved, we can host our BI (business intelligence) environment also within Snowflake. In our case, to provide basic reports and dashboards, we started to use Tableau, Power BI, Looker, and Qliksense, depending on our customer preference.

We are yet to figure out how to integrate tools, such as Liquibase, to release changes to our data warehouse model. If Snowflake could guide us with some easy-to-use integration (similar to DBT integration), that would be great.

I've used the solution since 2020.

Stable

Scalable

Support can be enabled in the Snowflake UI.

In the past, we used Google Cloud SQL. However, Snowflake offered cost optimization among the many other useful features. They also introduced app building on top of the data hosted.

The initial setup is not difficult. Google Search will lead us to articles that can guide us on the setup of users, roles, warehouses, and access controls.

We did the initial setup on our own, and it was not difficult.

We constantly monitor the usage with grafana dashboards to keep the ROI growing and to assist/ alert users about any wastage.

Many interesting features are available only in the enterprise edition. Check out the differences when you are evaluating the product: https://docs.snowflake.com/en/...

We considered MySQL and Google Big Query. We're also happy with Google Big Query.

Snowflake is growing with newer features and capabilities. But not much success with Stream lit app. Big query + app sheet is an alternative that we're considering.