AI Customer Data Analysis Automation

I want to take the customer data we already collect—from our CRM, support tickets, and purchase history—and run it through a fully automated, AI-driven analysis pipeline. The end goal is a repeatable process that cleans the raw data, surfaces patterns and actionable insights, and delivers them in clear visual or written reports without manual touch-points each time new data lands.

Here is what I have in mind:

• Data ingestion & cleaning: pull the existing customer tables (CSV exports and a live PostgreSQL instance), handle missing fields, and standardise formats automatically.
• Model layer: apply the right mix of statistical analysis and machine-learning techniques to uncover trends, cohort behaviours, and predictive indicators of value or risk.
• Insight delivery: generate a scheduled report—dashboard, PDF, or both—so non-technical stakeholders can grasp the findings at a glance.

I’m comfortable with common stacks such as Python, SQL, pandas, scikit-learn, and a lightweight BI tool like Metabase or Power BI, but I’m open to your recommendations if they improve speed or accuracy.

Acceptance criteria
– A script or workflow I can run on a fresh machine with minimal setup (README included).
– At least one automatically generated report demonstrating the analysis on our sample dataset.
– Clear commentary in the code so we can extend or tweak the models later.

If this sounds like your kind of project, let me know how you would tackle it and any libraries or cloud services you’d lean on to keep the solution robust yet cost-efficient.