Skyflow for Snowflake UDF Setup Instructions

Pre-requisites:

• A Snowflake account with Account Admin privileges
• A Skyflow account with Account Admin privileges - Contact Skyflow here to get one!
• A command-line terminal with python installed

Setup Steps:

1. Clone (or Download+Unzip) this repo, and navigate to the snowflake_udf_demo folder in a command-line terminal.

2. Log into Skyflow via browser and create a service account. Ensure the Account Admin role is enabled when going through the creation wizard. Move or copy the auto-downloaded credentials.json file into the snowflake_udf_demo folder.

3. Within Skyflow, click the settings gear icon on the left. Copy your Account ID and paste it into the vault_details.cfg file.

4. In your terminal window, run the command: python script_setup.py or python3 script_setup.py (depending on your python version). This

5. Log into Snowflake and click Create Worksheet from SQL File in the top-right three-dot menu. Select and Import the skyflow_udf_setup.sql file. Repeat for the skyflow_udf_demo.sql file.

6. Open the imported skyflow_udf_setup.sql worksheet and ensure you are using the Snowflake ACCOUNT_ADMIN role. Execute the worksheet in full using the Run All option in the Run command button drop-down (top-right). If you receive the message Function SKYFLOW_PROCESS_PII successfully created., the setup is successful!

7. Switch to the imported skyflow_udf_demo.sql worksheet, and again ensure you are using the Snowflake ACCOUNT_ADMIN role. Proceed line by line through the demo script. Prior to running the tokenize table step, update the last parameter with your email address. Note that this email address must correspond to an existing Skyflow Studio user in the relevant account. Also: Ensure that vault name remains "SkyflowVault".

8. Upon running the tokenize table step, you can refresh your Skyflow Studio web page - there will be a new vault and table in Skyflow securely hosuing your Snowflake PII data! Continue through the demo steps accordingly.

--This concludes the demo. To reset the Snowflake portion of the demo, import the skyflow_udf_reset.sql file as a worksheet and execute. You can then start again from the Run Demo portion of these instructions. Thank you!--