PUBLIC UNIVERSITY
BUSINESS PROBLEM
This prestigious Public University, known for its diverse student population and commitment to excellent student services, developed an internal chatbot tool to support preservice teachers—students pursuing teaching certification. The chatbot provides an environment for these students to practice their teaching skills. However, the development team faced challenges with migrating, scaling, and optimizing the chatbot’s architecture. To address these issues, they turned to Scalesology’s technical expertise and sought to leverage cloud technology to enhance the chatbot’s performance and scalability.
SCALESOLOGY IN ACTION
In the first phase of this project, Scalesology focused on migrating the application to a cloud environment. While the locally hosted app was sufficient for a small number of users, it could not support the demands of a large and growing user base. Scalesology ported the application, including all its services, to the Amazon Web Services (AWS) cloud infrastructure. This initial phase involved running the application services on a single server as a preliminary prototype in the cloud. The plan was to later scale the application to run on multiple servers, with each service hosted on its own dedicated server.
In the second phase, Scalesology addressed the scaling and optimization of the chatbot application. Once the app was running on the AWS cloud infrastructure, an architectural bottleneck was identified, causing delayed chatbot responses. This issue arose from the in-memory database sharing the same server as the resource-intensive large language model powering the chatbot, which degraded performance. Scalesology resolved this by migrating the in-memory database from the EC2 instance to Amazon ElastiCache. This migration significantly improved the speed and overall performance of the chatbot system in the cloud.
RESULT
The migration, scaling, and optimization of the internal chatbot application allowed a large user base to access and utilize the system seamlessly. The enhanced scalability ensured the chatbot could handle a higher volume of simultaneous inquiries, significantly reducing response times and improving student satisfaction. Additionally, the university benefited from reduced IT management and maintenance overhead, as the cloud provider handled infrastructure and security updates. By adopting a pay-as-you-go model, the university also achieved cost savings compared to maintaining and scaling on-premises servers.
SERVICE REFERENCES
Commentaires