From NAS to AWS S3: Storing Documents and Metadata with Fast2

In the previous episodes of our document migration series, we retrieved business files from a file system, interpreted folder paths as business properties, and enriched documents with structured metadata. Now, it’s time to take the next big step: storing both the content and its metadata in AWS S3, using Fast2.

This article shows how to automate the process of uploading documents and storing metadata as JSON files in S3 buckets, setting the stage for scalable, cloud-native ECM strategies.

Why Use AWS S3 for Document and Metadata Storage?

Amazon S3 offers secure, scalable, and highly durable storage — making it an ideal destination for enterprise content migration.
Storing documents in S3 along with structured metadata (in JSON format) allows for:

• Easy retrieval and indexing
• Compatibility with AI tools for enrichment or validation
• A future-proof, cloud-native architecture

With Fast2, this process can be automated and modular — no need for manual uploads or fragile scripts.

Step-by-Step: Automating the S3 Upload Workflow with Fast2

1. Upload Documents to AWS S3

We start by using the AWS injector module in Fast2 to upload the binary document content.

Configuration steps:

• Select the target S3 bucket
• Provide credentials (access key, secret key, region)
• Define the storage path and filename using the file name metadata field

Once executed, the files are uploaded and Fast2 stores the S3 key as migration-oriented data, useful for future reference.

2. Clean Metadata Before Storing

We don’t need to store all metadata — only what matters.

To streamline the JSON output:

• Use a JavaScript task to define a whitelist of metadata to retain
• Remove any metadata not included in the whitelist
• Clear folder references no longer needed
  

This ensures the JSON file will be concise and relevant.

3. Update Document Content Reference to S3

• Use a task to attach the new content path based on the stored S3 key
• Remove the old local path reference
• Disable content existence checks (Fast2 won’t validate S3 paths directly)

This ensures the migration item accurately reflects the actual content location.

4. Write Metadata to JSON Files

Next, we serialize the cleaned metadata into JSON format:

• Use the appropriate Fast2 task to output metadata as .json files
• By default, these are stored in: files/[campaign_name]/[task_name]/[document_id].json

This structure ensures organized and traceable metadata storage, ready for downstream processing.

5. Attach the JSON Metadata File to the Document

The generated JSON metadata file is then attached to the document as side content:

• Use a task to attach content
• Enable "check if file exists" to avoid missing references
• Replace any outdated side content

This links the document with its metadata in a clean, manageable way.

6. Push the JSON Metadata to AWS S3

Finally, we push the JSON file to S3, using the same injector module:

• Adjust the file name to follow this format: [document_id]-properties.json
• Reuse existing credentials and bucket settings

This ensures each document in S3 is accompanied by its corresponding metadata file.

Results: A Clean, Structured Migration to S3

At the end of this workflow:

• Documents are stored in Amazon S3
• Metadata is cleaned, structured, and saved in JSON format
• All references are updated inside Fast2

This setup supports scalable ECM migrations, enables AI readiness, and future-proofs your content architecture.

🎬 Want to see it in action? Watch the video : 

What’s Next? AI-Based Metadata Enrichment

With documents and metadata safely in AWS S3, the next step is AI enrichment.

In the next episode, we’ll explore how to use content analysis to validate or enhance metadata — going beyond structure to add real business value.

Useful Resources:

• 🔗 Fast2 Official Website
• 🔗 Fast2 Documentation
• 🔗 AWS S3 Overview
• 🔗 More Fast2 Tutorials

💡 Questions or ideas? Leave a comment under the video, and let's discuss the best practices for optimizing your own migrations!