Project Overview

This project demonstrates how AWS cloud and AI services can be combined to build a serverless, event-driven text-to-speech automation pipeline.

The solution automatically detects text files uploaded to an Amazon S3 bucket, processes the text using Amazon Polly, converts the content into speech, and stores the resulting audio file in a separate S3 bucket.

The project demonstrates how cloud services can be combined to create scalable automation workflows without requiring continuously running servers.

Architecture

Text File → Amazon S3 → AWS Lambda → Amazon Polly → MP3 Audio → Amazon S3

When a .txt file is uploaded to the source S3 bucket:

1. Amazon S3 generates an object-created event.
2. The event triggers an AWS Lambda function.
3. Lambda retrieves the text file from S3.
4. Amazon Polly converts the text into speech.
5. The generated audio is saved as an MP3 file.
6. The MP3 file is stored in the destination S3 bucket.

This creates an automated serverless workflow from text input to audio output.

Key Capabilities

• Event-driven automation
• Serverless application architecture
• Automated text-to-speech processing
• Cloud-based file processing
• AI service integration
• Secure AWS resource access
• Automated storage and processing workflows
• Logging and error handling

AWS Services & Technologies

AWS Services

• Amazon S3
• AWS Lambda
• Amazon Polly
• AWS IAM
• Amazon CloudWatch / Lambda logging

Development

• Python
• Boto3
• AWS SDK

Security

• IAM policies
• IAM execution roles
• Least-privilege permissions
• Controlled access to S3 resources

Security Architecture

The Lambda function uses an IAM execution role with permissions required to:

• Read objects from the source S3 bucket
• Write processed audio files to the destination S3 bucket
• Use Amazon Polly for speech synthesis
• Write execution logs

The project demonstrates the use of role-based access and controlled permissions rather than embedding AWS credentials directly into application code.

Automation Workflow

The workflow is designed to operate automatically without manual intervention.

Upload → Trigger → Process → Convert → Store

This approach can be extended to support larger content-processing workflows and other automated media or document-processing scenarios.

Potential Applications

The architecture demonstrated in this project can serve as a foundation for solutions such as:

• Voice-enabled applications
• Accessibility tools
• Automated audio generation
• E-learning content processing
• Content publishing workflows
• Automated document-to-audio conversion
• Voice assistant components
• Customer communication systems
• AI-powered content processing

Skills Demonstrated

AWS Cloud Engineering · Serverless Architecture · AI Service Integration · Event-Driven Architecture · AWS Lambda · Amazon S3 · Amazon Polly · IAM · Python · Boto3 · Cloud Automation · Security · Logging & Monitoring

Project Type

Internal / Demonstration Cloud & AI Engineering Project

Completed

November 21, 2022

Project Repository

GitHub:
https://github.com/donval/My-top-repositories/tree/main

Project Outcome

The project demonstrates how AWS services can be combined to create a fully automated, serverless AI-enabled processing workflow.

It provides a practical example of integrating cloud storage, event-driven computing, AI services, identity and access management, and application logic into a single automated solution.

The architecture can be adapted and extended for broader AI automation, intelligent content processing, and cloud-native application use cases.