Integrate Cloud Storage for Raw and Output Datasets

[rahibAmin84]

Hook up cloud storage for raw data sets and output data sets

[rahibAmin84]

Description:
Currently, the project uses local data files for both input (raw) datasets and output (predicted ROI) results. To improve scalability, reproducibility, and collaboration, we need to integrate cloud storage. This will allow the system to:
1. Store raw input datasets centrally so multiple users or environments can access them.
2. Save model outputs (predicted ROI and posterior distributions) in a persistent, shareable location.
3. Facilitate future extensions, such as automated data updates or dynamic model retraining.

Proposed Requirements:
• Use a cloud storage service (AWS S3, GCP Cloud Storage, or Azure Blob Storage).
• Ensure secure access with API keys or IAM roles.
• Implement functions to:
• Upload raw datasets to the cloud.
• Download raw datasets into the model pipeline.
• Save output datasets (ROI predictions, uncertainty estimates) to the cloud.
• Maintain a clear folder structure,

[rahibAmin84]

Optional: Include logging of upload/download operations for reproducibility.

[achankra]

This is an awesome improvement that could be significantly more practical