Transforming Student Support Services for Riara University through an AWS-native Generative AI Knowledge Assistant
About the Customer
Riara University is a premier private higher education institution based in Nairobi, Kenya, operating within the academic sector. Recognized for its commitment to academic excellence, innovation, and student-centered learning, the university offers a wide range of undergraduate and postgraduate programs. Operating within the Small-to-Medium Business (SMB) market segment, Riara University relies heavily on modern digital platforms to coordinate administrative processes and academic delivery for its growing domestic and international student population.
Customer Challenge
Prior to the intervention, Riara University faced significant operational bottlenecks in managing student support, particularly regarding high-frequency, repetitive inquiries related to admissions, fee structures, registration timetables, and academic calendars. Over 60% of daily incoming queries consisted of identical FAQs, which were handled manually by administrative staff via emails, phone calls, and walk-ins. This manual approach resulted in severe response delays—often taking 12 to 24 hours—and created an unsustainable operational workload for administrative teams, especially during peak registration and admission cycles. Without an intelligent, automated solution, the university faced immediate risks of student dissatisfaction, communication gaps during critical academic timelines, and escalating administrative overhead. In the long term, these inefficiencies limited the university’s scalability and threatened its competitive edge in a modern, digitally driven higher education market.
Partner Solution
Riara University is a premier private higher education institution based in Nairobi, Kenya, operating within the academic sector. Recognized for its commitment to academic excellence, innovation, and student-centered learning, the university offers a wide range of undergraduate and postgraduate programs. Operating within the Small-to-Medium Business (SMB) market segment, Riara University relies heavily on modern digital platforms to coordinate administrative processes and academic delivery for its growing domestic and international student population.
The application is hosted on right-sized Amazon EC2 instances within a secure, private subnet of an Amazon VPC, with public traffic safely routed through AWS Web Application Firewall (WAF) and distributed via an Elastic Load Balancer (ELB). To optimize performance and reduce backend LLM latency, we integrated Amazon ElastiCache (Redis) to cache high-frequency session data and queries, while Amazon SQS was implemented to manage asynchronous background processes. The entire infrastructure was provisioned programmatically using AWS CloudFormation to guarantee consistent, repeatable deployments free of manual console interventions.
90%
Improvement in support responsiveness.
70%
FAQs in academics and administration automated
50%
Reduction in manual support workload
Results and Benefits
The production deployment of the Generative AI chatbot yielded immediate, quantifiable improvements across Riara University’s support landscape:
- Near Real-Time Response Delivery: Average student inquiry response times plummeted from a manual baseline of 12–24 hours to 1–2 seconds, representing a greater than 90% improvement in support responsiveness.
- Significant Support Automation: The chatbot successfully automated 60% to 70% of all frequently asked academic and administrative questions, resolving them instantly without requiring human intervention.
- Reduced Administrative Workload: The automation of repetitive queries directly translated to a 50% reduction in manual support workload for administrative staff, allowing team members to focus on high-value student advisory tasks.
- Continuous Operations and Optimization: By leveraging AWS-managed services, the university eliminated the need for expensive, dedicated GPU hardware, achieving a highly sustainable, cost-optimized, and secure production workload.