This project analyzes customer reviews from the Sephora product dataset using Snowflake Cortex LLM functions to extract sentiment insights at scale. It demonstrates a modern data stack pipeline - from infrastructure as code with Terraform, to transformation and modeling with dbt, all running on Snowflake.

Goal: Use Snowflake Cortex's LLM capabilities to perform sentiment analysis on customer reviews.

Dataset: Kaggle Sephora product reviews dataset.

Stack:

• Infrastructure: Terraform
• Cloud Data Platform: Snowflake
• Transformation: dbt
• LLM Analysis: Snowflake Cortex function - COMPLETE()

Features:

• Provision Snowflake objects (warehouse, databases, schemas, roles) using Terraform
• Load and structure Sephora reviews into Snowflake
• Transform raw data using dbt models
• Apply Cortex LLM functions to summarize customer sentiment
• Enable downstream usage: dashboards, product team insights, or ML training