Our journey to Snowflake monitoring mastery

We’re big fans of Snowflake, and for good reason. Its ability to enable teams across Canva is a game-changer, mainly because SQL is such a prevalent skill set today. Snowflake’s user-friendly data management approach lets us extend its usage far beyond the typical data warehousing tasks. The ongoing enhancements and new features Snowflake continues to roll out are pushing the boundaries of what’s possible, allowing us to explore use cases unlocked with functionality such as Streamlit and Declarative Pipelines (Dynamic Tables).

The platform has become deeply integrated into our daily operations and is used extensively across Canva. It’s not just the engineering teams that benefit from Snowflake; designers, product managers, and even sales teams leverage it to drive insights and improve decision-making. In fact, over two-thirds of Canva interacts with Snowflake in some capacity, whether through direct querying, using data-powered dashboards, or indirectly through downstream tools that draw from our data warehouse.

We designed our data infrastructure to be efficient and scalable. We ingest first-party data into Snowflake through Amazon S3, ensuring a seamless pipeline from our internal systems. For third-party data, we rely on Fivetran to handle the heavy lifting of ETL (Extract, Transform, Load) processes. This data flows into Snowflake and is exposed through external tables. From there, dbt (data build tool) steps in, orchestrating our data transformations to ensure that raw data is shaped into a format that supports analysis and reporting.

Leave a Comment