This guide demonstrates how to build an end-to-end AI-powered analytics platform for streaming media advertising using Snowflake. You'll deploy a complete solution that transforms raw subscriber data into actionable audience insights using Dynamic Tables, ML-powered predictions, Semantic Views, and Cortex Analyst.

The solution uses synthetic data representing a fictional UK-based streaming media company. All subscriber profiles, demographics, clickstream events, and ad interactions are generated for demonstration purposes only.

All data in this solution is synthetic and generated for demonstration purposes:

Data flows through four layers:

In this step, you'll create all the Snowflake objects needed for the Audience Analytics solution.

1. In Snowsight, click Projects, then Workspaces in the left navigation
2. Click + Add new to create a new Workspace
3. Click SQL File to create a new SQL file
4. Copy the setup script from setup.sql and paste it into your SQL file

Run the setup script to create:

The script automatically:

1. Loads pre-generated synthetic data from S3 into the INGEST and DATA_SHARING schemas
2. Transforms data through Harmonized layer with enrichment and aggregation
3. Creates ML features for LTV and churn scoring
4. Deploys four Streamlit applications
5. Creates a Semantic View and Snowflake Intelligence Agent

The Dashboard is your executive analytics hub providing subscriber insights, ad performance metrics, and behavioral analytics.

Location: AME_AD_SALES_DEMO.APPS.DASHBOARD

1. In Snowsight, navigate to Projects → Streamlit
2. Find DASHBOARD in the AME_AD_SALES_DEMO.APPS schema
3. Click to launch

This view lets you search and filter the subscriber base to find specific profiles.

How to Use It:

1. Search for subscribers - Type a name, email, or persona keyword in the search box. The table updates as you type.

2. Filter by Subscription Tier - Use the dropdown to show only subscribers on a specific plan:

   • Ad-supported - Free tier with ads
   • Standard - Paid tier
   • Premium - Top tier

3. Filter by Persona - Narrow results to behavioral segments:

   • SPORTS_ENGAGED
   • PREMIUM_DRAMA
   • FAMILY_HOUSEHOLD
   • KNOWLEDGE_SEEKER
   • LIFESTYLE_MINDED
   • REALITY_FAN

4. Drill into a profile - Click any Profile ID. A detail panel shows LTV score, churn probability, engagement metrics, and demographic details.

This view shows advertising campaign metrics over time.

How to Use It:

1. Select a Campaign - Use the Campaign ID dropdown to focus on a specific campaign, or leave it on "All" for aggregate performance.

2. Filter by Content Category - Choose content genres: Sports Enthusiasts, Premium Drama, Family Entertainment, Comedy Nights, Documentary & Knowledge, Lifestyle & Wellness

3. Adjust the date range - Use date pickers to zoom into a specific time window.

4. Read the charts - Top chart shows eCPM (effective cost per thousand impressions) trends. Bottom chart shows CTR (click-through rate) over time. Hover for exact values.

This view analyzes user behavior through two sub-tabs.

Clickstream Events Tab:

1. Filter by Persona or Subscription Tier using the dropdowns
2. The heatmap shows event intensity by hour and day (darker = more activity)
3. Use this to identify peak engagement windows for your audience

Ad Events Tab:

1. Filter by Campaign ID or Target Persona
2. The chart shows daily metrics: Impressions, Clicks, and Completions (video ads watched fully)
3. Compare campaign performance across different audience segments

This view provides a conversational AI interface powered by Cortex Analyst. Ask natural language questions about your subscriber data and get instant SQL-backed answers.

How to Use It:

1. Select Analytics Assistant from the sidebar
2. Click any sample question to get started, or type your own in the input box
3. The assistant translates your question into SQL, runs it, and returns results with charts and tables

Sample Questions to Try:

The Ingest Explorer shows you the raw data before any transformation—this is where the data journey begins.

Location: AME_AD_SALES_DEMO.APPS.INGEST_EXPLORER

1. In Snowsight, navigate to Projects → Streamlit
2. Find INGEST_EXPLORER in the AME_AD_SALES_DEMO.APPS schema
3. Click to launch

View Your Data Sources: The top section shows connected stages (external data sources), indicating where raw data originates.

Browse Available Tables: The middle section lists all tables in the INGEST schema with row counts:

Inspect a Table: Click any table name to see:

1. Columns panel - Column names and data types (VARCHAR, NUMBER, TIMESTAMP, etc.)
2. Statistics panel - Min/max values, averages, distinct value counts, null counts
3. Sample values - Actual values for columns with few unique values
4. Data preview - First 50 rows to eyeball the data

The Dataset Explorer shows how raw data transforms into analytics-ready tables that power the Dashboard.

Location: AME_AD_SALES_DEMO.APPS.DATASET_EXPLORER

1. In Snowsight, navigate to Projects → Streamlit
2. Find DATASET_EXPLORER in the AME_AD_SALES_DEMO.APPS schema
3. Click to launch

Select a Schema: Use the sidebar to explore the data flow layers:

Explore the Lineage Graph: The interactive diagram shows how tables connect:

Profile a Table: Click any table to see column names and types, value distributions, and sample data.

The Segment Builder shows all attributes available for building audience segments—this is how you turn data into action.

Location: AME_AD_SALES_DEMO.APPS.SEGMENT_BUILDER

1. In Snowsight, navigate to Projects → Streamlit
2. Find SEGMENT_BUILDER in the AME_AD_SALES_DEMO.APPS schema
3. Click to launch

The left panel organizes attributes into categories:

Add Conditions: Click the "Add" button next to any attribute. It appears on your segment canvas.

Configure Each Condition:

Combine with Logic:

Click "Add subgroup" for nested logic.

High-Value Subscribers at Risk of Churning:

Young Adults with High Digital Engagement:

In Snowsight, click AI & ML, then Snowflake Intelligence in the left navigation.

The setup script automatically creates the Semantic View and Snowflake Intelligence Agent:

Semantic View (AME_AD_SALES_SEMANTIC_VIEW):

Snowflake Intelligence Agent (SUBSCRIBER_ANALYTICS_AGENT):

Try these persona-specific queries:

Data Analyst:

Marketing Manager:

Ad Sales Executive:

Executive:

Congratulations! You have successfully built a comprehensive Media & Entertainment audience analytics solution with Snowflake Intelligence.