Snowflake data warehouse setup and SQL analytics on EV charging network data, with resource monitors, role-based access control, and Cortex AI functions for sentiment analysis and text classification.
This project analyzes an electric vehicle charging network using Snowflake. It covers environment setup, query governance, role-based access control, and AI-powered analysis of multilingual customer feedback.
- Environment Setup – Separate loading and query warehouses, S3 stage, and table creation with bulk data loading
- Resource Monitors – Monthly credit quota with automated notification and suspension triggers
- Statement Timeouts – Warehouse-level query timeout (60 min) and queue timeout (15 min) configuration
- Role-Based Access Control – Custom
charging_analyst_rolewith scoped warehouse, database, schema, and table privileges - SQL Analytics – Station stats, cumulative installations by year, session metrics, and feedback analysis using CTEs and window functions
- Cortex AI Functions –
AI_TRANSLATE,AI_SENTIMENT, andAI_CLASSIFYapplied to customer feedback for multilingual sentiment and topic classification
| Table | Records | Description |
|---|---|---|
stations |
150 | Charging station locations, bay counts, and power ratings |
charging |
35,156 | EV charging sessions with start/end times and energy delivered |
user |
1,166 | Customer feedback with ratings and multilingual comments |
- 150 stations with 718 total bays and 18,800 kW combined power capacity
- 35,156 charging sessions across 2024; average of 9.05 sessions per user
- 66% of occupied time was active charging; 34% (949,121 minutes) was idle
- Average customer rating of 3.21/5; feedback frequently cites idle congestion as the top issue
- AI classification and sentiment analysis support implementing an idle fee policy to improve bay turnover
Snowflake · SQL · Snowflake Cortex AI · AWS S3