feature(ai-tutorials): add snowflake cortex livestock weather intelligence connector

[kellykohlleffel]

Summary

Demonstrates Snowflake Cortex Agent integration for real-time AI enrichment in Fivetran data pipelines. This pattern can be applied to any industry vertical for intelligent data augmentation during ingestion.

What This Connector Does

• Real-time AI enrichment via Snowflake Cortex Agent REST API (pattern applicable to any industry vertical)
  • Livestock health risk assessment based on weather conditions
  • Farm-specific actionable recommendations
  • Species-specific risk analysis
  • Historical weather-health correlations
• Syncs weather forecasts for farm ZIP codes from NWS API
• Supports multiple ZIP codes with farm mapping
• Configurable enrichment limits for cost control
• Comprehensive error handling and retry logic

APIs Integrated

1. Snowflake Cortex Agent - AI enrichment with llama3.3-70b and Cortex Analyst
2. National Weather Service - 7-day weather forecasts
3. Zippopotam.us - ZIP code to coordinates conversion

Key Features

• ✅ Snowflake Cortex Agent REST API integration with real-time enrichment during ingestion
• ✅ Cortex Analyst semantic views for intelligent data querying
• ✅ Configurable enrichment limits for cost control
• ✅ Comprehensive error handling with exponential backoff
• ✅ Complete setup guide for Cortex Agent and semantic views
• ✅ Sample AGR_RECORDS data (42 records, 4 farms)
• ✅ Unit tests included (18 test cases)

Use Cases

• Integrating Snowflake Intelligence and Cortex Agents into Fivetran real-time pipelines for enrichment
• Proactive livestock health management with AI-driven insights
• Emergency weather preparedness with predictive analysis
• Historical weather-health correlation analysis
• Insurance documentation and regulatory compliance

Tags

#snowflake-cortex-agent #snowflake-intelligence #ai-enrichment #real-time-enrichment #weather #agriculture #livestock

Jira Ticket

N/A

Description of Change

New quickstart example demonstrating Snowflake Cortex Agent integration for real-time AI enrichment during Fivetran data ingestion. Provides reusable pattern for any industry vertical requiring intelligent data augmentation.

Testing

• ✅ Tested locally with fivetran debug - all syncs successful
• ✅ Successfully syncs weather data with AI enrichments
• ✅ Handles API failures gracefully with retry logic
• ✅ Unit tests pass (18/18 tests)
• ✅ No credentials or PII in submission
• ✅ All example files use placeholders only

Checklist

• Tested the connector with fivetran debug command
• Added/Updated example-specific README.md file with complete documentation
• Followed Python Coding Standards (module-level constants, PEP 257 docstrings, complexity < 10)