How Dreamcast Optimized Event Image Delivery with AWS & Thumbor for 60% Faster Load Times

Company Overview

Dreamcast is a top event-tech firm that provides end-to-end solutions for virtual, hybrid, and in-person events. Focused on innovation and hassle-free event experiences, Dreamcast provides a strong portfolio of services such as customizable virtual event platforms, high-definition live streaming, on-site RFID and access management, online ticketing, and interactive audience engagement solutions.

With services to clients in India and internationally, Dreamcast supports various industries like education, corporate, healthcare, entertainment, and government. From virtual seminars to hybrid conferences, product launches to large-scale physical events, Dreamcast helps organizations provide effective and lasting experiences.

What distinguishes Dreamcast is its capacity to align the latest technology with client-specific requirements for high engagement, scalability, and seamless execution. Driven by a strong focus on quality, innovation, and client satisfaction, Dreamcast has become a go-to partner for organizations that seek to transform the event experience in a digital-first era.

Business Requirement

With Dreamcast’s new offerings to support mass-scale virtual expositions and webinars, it faced major technical issues affecting user experience and platform performance.

1. Performance & Latency

Customers across geographies saw slow loading of images on event landing pages. The absence of dynamic optimization of images based on device type or network speed resulted in variable performances across regions and devices.

2. Scalability

With hundreds of thousands of live attendees participating at the same time, Dreamcast found it difficult to grow its image hosting infrastructure. Manual image processing workflows that existed were not suitable for handling spurts in traffic, which resulted in lag and poor user experiences.

3. Technical Modernization

The use of legacy, static image processing scripts caused a bottleneck. Dreamcast sought to bring this system up to date by implementing **Thumbor**, an open-source, containerized image processing service. This would allow for advanced features like smart cropping, filters, and dynamic resizing—built directly into their event platform.

4. Operational Agility

To back real-time events, Dreamcast needed a containerized deployment strategy that enables instant scaling without service disruption. They also wanted secure, automated CI/CD pipelines to deploy, update, and manage image assets efficiently while ensuring reliability and velocity.

Architecture Solution

To solve the scalability and performance issues, Dreamcast partnered with Cloudminister to create a scalable, cloud-native architecture on AWS. The solution involved containerization, serverless deployment, and worldwide content delivery for maximum efficiency.

1. Docker Image Creation

The application for the event—involved in image processing and delivery—was containerized with Docker.

Key capabilities were:

• Global access to images
• Instantaneous transformation of images
• Smooth integration with AWS CloudFront

2. Amazon ECR (Elastic Container Registry)

The Docker image was hosted in ECR, allowing secure access, version management, and seamless deployment to AWS resources.

3. AWS App Runner:

App Runner pulled the Docker image from ECR and ran it with auto-scaling, HTTPS built-in, and load balancing—providing high availability during event bursts.

4. Amazon S3 + CloudFront Integration:

S3 served as the main image storage.

CloudFront was setup to:

• Cache content globally at edge locations
• Deliver static images or route dynamic requests to App Runner
• invalidate cached images in real time that are no longer current

End-to-End Workflow

Event images were accessed by attendees through CloudFront URLs. Depending on the request:

• Cached images were instantly returned
• Uncached requests invoked App Runner for dynamic processing
• Processed images were cached back to S3 to be re-used for subsequent requests

Action Taken

1. Containerized Image Services – Compressed the image upload and optimization features into Docker containers to enable consistent and portable deployment.
2. CI/CD with GitHub Actions – Configured automated pipelines to create, test, and release the containerized application, promoting quicker and more trustworthy releases.
3. S3 Lifecycle Policies & Versioning – Activated lifecycle rules to automatically handle image storage and cleanup, while versioning provided rollback features and data integrity.
4. CloudFront Invalidation Configuration – Incorporated real-time cache invalidation so the latest image updates were immediately visible in all the edge locations.
5. Performance Monitoring – Implemented AWS CloudWatch and AWS X-Ray to monitor system health, API performance, and diagnose latency or bottlenecks with ease.

Outcome Highlights

• 60% Quicker Image Load Time — Attained via optimized delivery practices.
• Reduced Global Latency — Utilized Amazon CloudFront to deliver content quickly across the globe.
• Improved Scalability — Used AWS App Runner for auto-scaling based on user demand.
• Decreased DevOps Overhead — Simplified deployment and infrastructure management.
• Increased Reliability — Provided high uptime and fault tolerance for events hosting 50,000+ users.