Technical Projects

This project automated weekly vendor reporting by collecting e-commerce report data, converting raw JSON files into structured CSV reports, storing both raw and processed files in Amazon S3, and distributing scheduled report emails through Amazon SES.

The workflow reduced manual report handling and created a scalable foundation for future reporting expansion across multiple e-commerce platforms.

Solution topology: automated e-commerce report workflow

1. Amazon EventBridge triggers the workflow on a weekly schedule.
2. AWS Lambda retrieves API credentials securely from AWS Secrets Manager.
3. The system authenticates with multiple e-commerce platforms and downloads reports in JSON format.
4. A processing Lambda function automatically converts JSON files into structured CSV reports.
5. Raw JSON files and processed CSV files are stored in Amazon S3 buckets.
6. Amazon SES sends report emails to different user groups.
7. Processed reports are automatically pushed to vendor FTP servers for integration with their internal systems.

• Automated weekly report generation and delivery.
• Stored raw JSON and processed CSV files in Amazon S3.
• Used DynamoDB to track report processing status.
• Delivered CSV reports through an SES email workflow.

• Amazon SP-API
• Walmart Marketplace
• Shopify

E-commerce and small or medium-sized business websites often need simple communication features, such as contact forms, inquiry forms, or quotation request forms. These tools support sales, customer service, and lead generation in a reliable and low-maintenance way.

A serverless contact form architecture on AWS is well suited for this use case. By using Amazon S3, CloudFront, API Gateway, Lambda, and SES, a business can build a low-cost, scalable, highly available contact workflow without managing traditional servers.

In this project, I implemented a serverless email function for a static website using multiple AWS services.

• Amazon S3 hosts the static website files, including HTML, CSS, and JavaScript.
• Amazon Route 53 manages DNS and custom domain routing.
• Amazon CloudFront delivers website content globally through a CDN.
• AWS Certificate Manager provides SSL/TLS certificates for secure HTTPS access.
• Amazon API Gateway receives frontend form submissions and routes them to the backend.
• AWS Lambda processes the contact form logic.
• Amazon SES sends the submitted message to the designated email account.

When a user submits information through the website contact form, JavaScript first handles the form data on the frontend and then sends the request to the backend through Amazon API Gateway. The request is processed by a Lambda function, which handles the business logic and triggers Amazon SES to send the message.

This architecture demonstrates how serverless cloud services can be used to build a scalable and efficient communication system for business websites.

Solution topology: AWS serverless contact form architecture

This project demonstrates how to design and deploy a personal cloud-based VPN gateway using WireGuard on AWS EC2, giving users full control over their data, encryption, and network path.

The VPN allows a laptop or mobile phone to connect securely to the internet through an encrypted tunnel, reducing exposure to risks on public networks.

The project also addresses a key operational question: how to manage VPN users after the VPN server is deployed. To solve this, I built an AWS-based Admin Portal to simplify user management and automate the provisioning process.

Solution topology: VPN server

1. Log in securely with Amazon Cognito.
2. Create and terminate VPN users.
3. Automatically generate WireGuard configuration files.
4. Store user profiles securely in Amazon S3.
5. Push VPN configuration updates to EC2 through AWS Systems Manager.
6. Send VPN access files to users through Amazon SES.
7. Manage access through a secure web-based admin portal.

The result is a cloud-based VPN management platform designed for secure user provisioning, automated configuration, and centralized administration.

Solution topology: VPN admin portal

This application uses a Retrieval-Augmented Generation architecture so chatbot answers are grounded in the uploaded PDF document.

The project allows users to upload a PDF, ask questions in natural language, receive AI-generated answers, and view source page references for better transparency.

• Upload a PDF document.
• Ask questions using natural language.
• Receive answers generated by an AI model.
• View source page references from the uploaded PDF.

This project demonstrates how a document-based chatbot can combine a user interface, a language model, an application framework, and a vector database to provide contextual answers based on uploaded content.

Solution topology: PDF chatbot RAG workflow