VUS builds connected English learning ecosystem with Google Cloud

VUS reception lobby, Photo courtesy of VUS

In many conventional learning environments, English study is largely confined to classroom hours. While students continue practicing at home, the process is often disconnected from teachers and parents, making it difficult to consistently monitor progress. Teachers typically review assignments only in the next class, while parents have limited visibility into their children's daily engagement.

To address this gap, VUS launched OVI, its digital learning ecosystem designed to create a continuous and connected learning journey. OVI stands for Outcomes-oriented and Opening up to online x offline, powered by VUS, with a focus on individualized learning for each student.

The platform enables students to practice language skills, review lessons, and access learning materials from home, while giving teachers and parents real-time access to progress tracking and performance data.

With more than 80,000 students across its system, VUS required a platform capable of handling large-scale concurrent usage while maintaining stable performance. Real-time quizzes, data synchronization, and video streaming all needed to operate smoothly, particularly during peak periods when traditional infrastructure can face delays or downtime.

"We had ambitious growth plans, but traditional infrastructure carried too much risk for us. We needed a serverless, autoscaling, and globally reliable backbone to support our vision," said Tan Nguyen, IT director at VUS.

Teacher and students during class, Photo courtesy of VUS

To build the system, VUS selected Google Cloud for its integrated ecosystem spanning data, AI, and application delivery. Three core services play a central role: BigQuery for data unification, Firebase for mobile-first learning experiences, and Vertex AI for AI-driven innovation.

VUS also partnered with Cloud Ace, which provided technical consultation and support in designing the OVI platform architecture.

The platform is designed for real-time responsiveness. During peak hours, student requests are distributed through Cloud Load Balancing across serverless Cloud Run services, enabling the system to scale from zero to more than 50,000 concurrent users without interruption.

For video lessons, Cloud CDN helps reduce latency by delivering content from edge locations closest to users. According to VUS, this architecture supports 99.9% uptime and has reduced application response time by 40%.

Teachers also gain real-time access to student data through Looker Studio dashboards powered by BigQuery. This allows them to identify students whose engagement or performance falls below expected thresholds and provide timely interventions, including additional tutoring when needed.

The transition to a serverless and automated system has also changed how VUS's internal teams operate. By using Cloud Build to automate its CI/CD pipeline, the time required to launch new features has been reduced from three days to three hours.

This faster development cycle enables teams to test new ideas and respond to student feedback more quickly. At the same time, Cloud Run's pay-per-use model has helped reduce monthly hosting costs by 25–30%.

Combined with a 60% reduction in infrastructure support tickets through Cloud Monitoring, the IT team has been able to shift its focus from system maintenance to improving learning analytics and enhancing the student experience.

VUS is now using these efficiencies to develop its next phase of innovation, centered on generative AI. Future plans include building AI-powered learning companions using Vertex AI and Gemini models, designed to provide 24/7 pronunciation feedback, encouragement, and personalized support.

Cloud Speech-to-Text is also expected to support automated English evaluations and speaking assessments.

"The biggest change is cultural. Building on Google Cloud helped us move from an operational-focused mindset to a data- and innovation-driven mindset," Tan said.