All Case Studies
Media & Entertainment U.S.A.

Building a Dynamic Augmented Reality App Without the Cloud Cost

Our client is an innovative tech startup focused on democratizing Augmented Reality. Their mission is to make AR content creation accessible to everyone, regardless of technical expertise. They have a strong vision for user-driven experiences. That's why they aimed to launch a dynamic platform where individuals and businesses could craft, customize, and share AR experiences using real-world visuals and objects. The client chose Brainium as their tech partner. They wanted a solution that would be cost-effective, scalable, and flexible enough to serve industries ranging from retail and education to entertainment and marketing. We took on the challenge with enthusiasm.

IndustryMedia & Entertainment
Tech StackKotlin, Swift, ARCore, ARKit, PHP, MySQL
CountryU.S.A.
01

The Challenges We Faced Along the Way

Most AR apps hard-code content during development, offering little flexibility for real-time changes. Dynamic content typically requires cloud-based recognition services, which are expensive, introduce privacy concerns, and often require sharing data with third parties. For startups and small businesses, this creates a barrier to entry due to high operational costs and vendor lock-in.

Moreover, not all user-uploaded images are suitable for AR marker recognition. Since the platform allows content creation by users worldwide, it needed a way to intelligently assess image suitability for AR usage without requiring expert intervention.

Another technical hurdle was enabling smooth video playback without relying on costly video streaming services. To keep the app affordable and scalable, the infrastructure needed to balance performance with cost-efficiency, and include systems for reliable backup and rapid recovery in the event of system failures.

Other challenges that we encountered include the following -

Dynamic AR Content

Making AR content dynamic and editable post-deployment without cloud services.

Image Scoring

Filtering and scoring user-submitted images to ensure AR marker compatibility.

Multimedia Playback

Supporting seamless multimedia playback without additional streaming infrastructure.

Cost-Efficient Backend

Maintaining a low-cost backend while ensuring data integrity, uptime, and responsiveness.

02

Our Primary Approach for This Project

We conducted a detailed consultation with the client. We took time to understand their needs and requirements. To deliver on all fronts, we shifted to building native Android and iOS apps using Kotlin and Swift, respectively, and their native AR frameworks-ARCore and ARKit. We opted for this since it would allow faster, smoother AR performance tailored to each platform while maintaining cost efficiency.

Furthermore, we decided to develop a custom marker evaluation system to rate image suitability and filter unsuitable ones. This would allow global users to freely generate AR content using real-world images, banners, posters, and even physical objects. Thanks to extensive optimization, we could guarantee smooth performance even on limited-resource mobile devices.

We also conferred with the client's team to build a backup system that distinguishes between frequently updated content and less volatile system components to balance cost and reliability. Our team came up with plans to develop a companion AR website with nearly all mobile features for broader access.

03

Our Primary Solution for This Project

Kotlin Swift ARCore ARKit PHP MySQL

To build a future-proof AR platform, we adopted an approach that minimized recurring costs and maximized device-native capabilities.

Runtime Marker Library Creation

Using Unity's AR Foundation in earlier versions and then switching to platform-native capabilities (ARCore for Android and ARKit for iOS), we enabled dynamic marker recognition at runtime. This eliminated the dependency on third-party cloud services for image recognition, significantly reducing operational costs and improving response times.

Native AR Engine Integration

By leveraging ARKit and ARCore directly through Swift and Kotlin, the app delivers faster recognition and smoother performance with fewer compatibility issues. This also allowed us to fine-tune performance for each OS, resulting in a better user experience.

Server-Side Marker Evaluation

We integrated an intelligent server-side image scoring system that evaluates every uploaded image's suitability as an AR marker. Images are assigned scores based on clarity, contrast, and uniqueness. If an image score is too low, the user is prompted to upload a better alternative.

Optimized Media Handling

All uploaded videos and images are compressed and encoded using parameters specifically tailored for the app's performance requirements. This ensures high-quality playback without requiring a streaming backend, making the system lighter and more responsive.

Dedicated Storage Architecture

AR content is stored on a high-throughput dedicated disk, while the core system files reside on a separate disk. The content disk is backed up frequently, and system backups are triggered only when major updates occur. This strategy balances data security and cost efficiency.

Cross-Platform AR Website

Alongside the mobile apps, we developed a responsive AR web portal with similar features excluding real-time scanning. It includes an auto-play feature for content, enabling users to view shared AR experiences via browser.

Cloud Infrastructure on AWS

The entire solution is hosted on AWS, using a suite of services configured for scale and reliability. Monitoring tools alert the team of any system anomalies, such as high traffic loads, low disk space, or server downtime, allowing quick response and uninterrupted service.

04

Results

Our solutions allowed our client to offer both businesses and users a unique, interactive, and scalable AR experience without relying on expensive third-party tools.

90% Lower Infra Costs

Eliminating cloud-based recognition led to substantial savings for the client and users.

Hear From Our Client Their Experience Partnering With Us

We loved working with Brainium on this project. The developers understood the assignment from the get-go and we didn't have to micromanage them at every turn. We're looking forward to future collaborations with the team.

Carrie Mahoney

Fulfilling Client Needs One Step At A Time

By blending device-native AR capabilities with a cost-efficient infrastructure, Brainium successfully delivered on every core client requirement. We eliminated cloud dependency, and enabled dynamic, user-generated content creation. The result is a highly scalable AR platform that empowers creativity, promotes user engagement, and adapts effortlessly across industries.

This project exemplifies how strategic use of native technology and thoughtful engineering can unlock the full potential of Augmented Reality. By empowering users to create immersive experiences from everyday visuals, we helped our client deliver an accessible, scalable, and high-performing AR platform ready to transform engagement across industries.

Want Results Like These?

Tell us about your project and we'll show you how to get there - faster than you think.

Leaving Already?
Let us help you find the right services for your business!

Our expert will help you in:

  • the right solution for your business
  • A ballpark estimate
  • An estimated delivery time

Start the Conversation!

Reach Out to Our Team