The Context: The Before State
We helped Language Academy fix a slow and unstable learning platform while rebuilding it for the future, without disrupting their day to day operations.
Language Academy is an Australia based education company focused on language learning. With a growing user base and increasing reliance on digital delivery, their website had become the core of their business.
For them, the platform was not just a website.
It was the classroom, the assessment system, and the customer experience all rolled into one.
The Clients Problem
Over time, the existing system began to show cracks.
Multiple bugs were affecting daily usage.
Several features were not behaving as expected.
API responses were slow, frustrating both users and internal teams.
On the backend, the application was executing duplicate database queries for common operations. This caused unnecessary load and poor performance at scale.
At the same time, the platform relied heavily on external AI APIs. These APIs were powerful, but expensive. Costs were rising quickly and replacing them with in house AI models was not realistic without compromising accuracy.
To make things more complex, Language Academy was planning a full website renovation. While the new platform was being built, the old one still needed to stay stable and usable.
The Goal
Success was defined very clearly.
Keep the existing platform running smoothly without disruption. Fix critical bugs and performance issues.
Reduce AI related operational costs. Deliver core learning modules for the new platform on time. Ensure the team could confidently launch before the event.
Anthea Strezze
Reliable partner with broad technical and design expertise!
A growing language learning platform struggled with bugs, slow API performance, rising AI costs, and a tight deadline for a complete website renovation.
A dedicated development team stabilized the existing platform, optimized backend performance, reduced third party AI dependency, and delivered core modules in parallel with the new website build.
40 percent performance improvement, 40 percent cost savings, 70 percent fewer bugs, and a stable platform ready for a major public facing event.
The Hardest Challenge
The platform was central to the business. It handled learning, assessments, and user experience across a growing user base.
Over time, the system began to show strain.
- Multiple bugs disrupted daily operations
- API response times slowed down user workflows
- Duplicate database queries increased system load
- AI API dependency caused rising and unpredictable costs
- System stability became unreliable under scale
At the same time, a full website rebuild was already planned. The problem was not just technical. The existing platform had to stay stable while the new one was being built. There was no room for disruption.
- Keep the existing platform stable without downtime
- Fix critical bugs affecting user experience
- Optimize backend performance and database efficiency
- Reduce AI related operational costs
- Deliver core modules for the new platform before the event
- Ensure confidence in launch readiness
The Solution
Instead of jumping into a full rewrite, the focus was on strengthening what already existed.
Execution followed a dual track approach.
One track stabilized and optimized the current platform
The other delivered core modules for the new website
Work was executed in controlled phases, ensuring every improvement was validated before moving forward.
The Outcome
The platform moved from instability to controlled performance.
- Performance improved across user workflows
- Bug frequency dropped significantly
- AI costs became predictable and controlled
- System downtime reduced noticeably
- Core modules were delivered on time
The UI team did an awesome job of understanding the functionality of our old site and the requirements we had for the new site. They were extremely knowledgeable about the options available in terms of technology, and made recommendations specific to our needs.
