snowpipe_streaming_dts_summit_demo-main

Snowflake Snowpipe Streaming + DTs Demo

This demo does the following:

• Records audio from the default microphone
• Does speech to text transcription
• Pipes those words to Snowflake using Snowpipe Streaming
• Has a DT that aggregates words and their counts
• Has a DT that uses the aforementioned DT with an ads table to determine what ads to show to the user

Snowflake Setup

Set up a role that has create database/warehouse, ex:

use role accountadmin;
create or replace role tjones_summit_rl;
grant create database on account to role tjones_summit_rl;
grant create warehouse on account to role tjones_summit_rl;
grant role tjones_summit_rl to user tjones;

Set up a private/public keypair for the user by following these instructions.

Create a profile.json with the following attributes:

• user: Login Name or Name of User
• url: The account URL, ex https://account.snowflakecomputing.com
• account: Name of the account
• private_key: The private key contents, without headers
• port: Port to use, use 443 for secure connections
• host: Account URL, ex account.snowflakecomputing.com
• role: The role to use, ex tjones_summit_rl

Create the following directories locally:

mkdir audio;
mkdir audio_raw;

Build the Client SDK driver:

mvn clean install;

Set the following environment variables:

• SNOW_ACCOUNT: The name of the account, ex account
• SNOW_ROLE: The name of the role to use, ex tjones_summit_rl
• SNOW_USER: Login Name or Name of User
• SNOW_PW: Password of the user.

Set up the Python virtual environment:

python -m venv venv;
source venv/bin/activate;
pip install -r requirements.txt

Run the demo, either via Python or the run_demo.sh script:

python summit_demo.py -nopython

OR

./run_demo.sh