mohitbagh01-rgb/Cloud-metrics-dbt ? reverse-engineered prompt

Build me a dbt analytics project for a fake SaaS company called CloudMetrics.

I want it to take raw customer, subscription, invoice, payment, session, feature, and feature usage data, then transform it into clean analytics ready tables in layers. Start with a staging layer that standardizes the raw data and removes anything unnecessary, then add reusable intermediate models for things like user sessions, invoice payments, organization subscriptions, and feature usage, and finally create business friendly marts for customer analytics, organization engagement, SaaS KPIs, subscription revenue, user engagement, and product usage.

Please include data quality tests like not null, unique, and relationship checks on the important keys and foreign keys, and make sure the project is documented so the model lineage and columns are easy to understand. It should be set up to work with Databricks tables and data coming from S3, and use dbt source definitions, refs, and YAML based configs. If you need to check anything, look up the current dbt docs online.