Loading use case index…
Loading use case index…
AI use case
University of Cape Town's EduBot served 43,600 students and 245,000 answers in 90 days post-AI upgrade, with 93% resolved without human help, 4.8/5 satisfaction, and 28% drop in…
Core facts from this catalog record. Primary narrative lives in the hero above; full raw fields follow in the next section.
Every column from the source row, in stable order. URLs open in a new tab.
Title
University of Cape Town EduBot: 43,600 Students Served in 90 Days
Content
The University of Cape Town (UCT) upgraded its rule-based EduBot with AI on 28 May 2025, transforming student support while keeping accuracy a top priority. Within 90 days of launch, EduBot engaged 43,600 unique users and delivered 245,000 AI-powered answers — an average of 19 messages per conversation. The system achieved a 4.8/5 star rating from 3,823 user ratings, with a 97% satisfaction score. "The AI upgrade has reduced routine questions so that staff can now deal with complex cases," says Dr Riashna Sithaldeen, project lead at UCT. The most dramatic operational shift was a 28% reduction in human handoff requests: the share of students requesting human assistance dropped from 11% before AI (Aug 2021 – 27 May 2025) to just 8% in the first 90 days post-launch. Live chat requests fell from 6% to 3.3%, and support ticket submissions decreased from 6% to 4.7%. Most remarkably, 93% of users had their questions answered without requiring human intervention. The motivation was a problem with UCT's first-generation, rule-based chatbot: it had processed over 6.8 million messages from 277,000 unique users with an 84% self-service resolution rate, but the analytics revealed that "I want to speak to a human" consistently ranked in the top five student queries. Students were growing frustrated with static, scripted responses and wanted natural conversation rather than menu navigation. Staff also faced scaling issues, constantly adding and maintaining questions. The architecture was a hybrid that built on UCT's proven rule-based system rather than replacing it. EduBot was trained on comprehensive institutional knowledge including official documents, student resources, and website content. Staff only needed to provide data sources and links, which the system processed into accurate responses. The multilingual system supports 65 detected languages and 100+ supported, enabling UCT's diverse international student body — including 6,000 international students — to access help in their preferred language. 59% of messages were sent outside regular office hours. User demographics: 28,372 prospective undergraduates (65% of users), 7,691 prospective postgraduates, 6,000 international students. "UCT's experience shows that AI does not need to compromise accuracy to deliver a better student experience," said Mitchell Hawley, Founder and Managing Director of Verge AI, the implementation partner. UCT is now exploring short courses for lifelong learners, faculty-wide AI chatbot pilots, and AI-powered staff support systems — building on the success of the admissions-focused AI chatbot to create a foundation for university-wide digital transformation.
Continue exploring AI deployments in the catalog.
Back to use casesCity
Cape Town
Company/Organization
University of Cape Town
Continent
Africa
Country
South Africa
Category
Diversified Consumer Services
Type
Deployment
Id
130e8fb7-6ac6-465a-961a-e64728f81dc5
Created At
2026-06-24T08:25:04.640267+00:00