Skip to content
Snowflake logoAmazon S3 logo

Connect Snowflake to Amazon S3

Move data between Snowflake and Amazon S3 the moment it happens. Build the flow visually, or let AI write it for you, then run it on our cloud or your own infrastructure.

Snowflake triggers
5
Snowflake actions
14
Amazon S3 triggers
1
Amazon S3 actions
8

Automation ideas

What you can build with Snowflake and Amazon S3

A few flows teams run between Snowflake and Amazon S3. Use one as a starting point, or build your own from scratch.

  • New Row in Snowflake creates Upload File in Amazon S3

    Triggers when a new row is inserted into the selected table. Uses a timestamp column you choose to detect rows added since the last check.

  • New or Updated File in Amazon S3 creates Run Query in Snowflake

    Triggers when you add or update a file in your bucket. The bucket/folder you choose must not contain more than 10,000 files.

  • New Row in Snowflake creates Read File in Amazon S3

    Triggers when a new row is inserted into the selected table. Uses a timestamp column you choose to detect rows added since the last check.

  • New or Updated File in Amazon S3 creates Run Multiple Queries in Snowflake

    Triggers when you add or update a file in your bucket. The bucket/folder you choose must not contain more than 10,000 files.

  • New Row in Snowflake creates Generate signed URL in Amazon S3

    Triggers when a new row is inserted into the selected table. Uses a timestamp column you choose to detect rows added since the last check.

  • New or Updated File in Amazon S3 creates Insert Row in Snowflake

    Triggers when you add or update a file in your bucket. The bucket/folder you choose must not contain more than 10,000 files.

Questions, answered

No. Pick a trigger on either piece and an action on the other, map the fields, and turn the flow on. Add a code step later only if you want to.

Yes. Build one flow that starts in Snowflake and acts in Amazon S3, and a second that runs the other way, in the same project.

On our EU cloud in Frankfurt, or self-hosted on your own infrastructure with Docker or Kubernetes. The MIT-licensed core means there is always an exit.