JOSH'S PORTFOLIO VAULT

Serverless Image Processing Pipeline

I built a fully serverless image processing pipeline on WS designed to automate the conversion of marketing images into standardised product-ready formats.

The application uses API Gateway and pre-signed URLs to securely upload batches of images directly to S3 without exposing the back-end. Upload events trigger Lambda functions that resize and convert images using a Pillow layer, storing processed outputs in a seperate bucket and generating a downloadable ZIP archive for retrieval.

The project focuses on event-driven architecture, asynchronous processing and secure file handling patters commonly used in production cloud systems. Through building and deploying the solution end to end, I gained hands-on experience with CORS configs, serverless workflows and scalable media processing pipelines.

F1 Data Driven Insights

A full design document to run through a project I did using data driven insights to benefit Formula 1 teams (and applicable to other racing industries).

Simple Static Website Hosting

A simple architecture to demonstrate static website hosting, similar to how this portfolio operates!

Basics of Identity and Access Management for Users

The core idea is that permissions don't live on Users - they live in policies.

An IAM User represents a person, like you or I. By default, we have zero permissions unless granted them via IAM.

Users are added to groups to simplify access management across teams, big or small. Policies can be attached to these groups, and automatically assigned to all users within that group. These policies define which actions are allowed or denied on specific AWS resources. If the policy allows it, users within a certain group can interact with services like EC2, S3 and many more.

IAM should always be implemented using the principle of least privilege, granting only the permissions required to perform a specific task.

AMAZON WEB SERVICES

Credential ID: 09ff41d6-54b5-42e5-9634-389ffefe0c08

Issued: September 2025

AMAZON WEB SERVICES

Credential ID: 69d9f1f2-1f8a-4a91-9169-b0b73414e8f7

Issued: January 2026

DISTRIBUTED COMMERCE SYSTEM

Advanced e-commerce architecture leveraging AWS serverless infrastructure. Implements SNS, SQS, Lambda, and DynamoDB for scalable order processing.

Serverless components handle high-throughput requests while Amazon SNS manages real-time notifications and automated fulfillment workflows.

PERSONALISED CLOTHING RECOMMENDATIONS

Machine learning application utilising AWS Rekognition for image analysis and SageMaker for predictive modeling. Integrates real-time weather data to generate contextual fashion recommendations.