evidoria

← Back to browse

Good practice Imported

MoUJi — University of Johannesburg's Multilingual AI Chatbot for Admissions, Registration and Financial Aid

South Africa · Johannesburg · See the South Africa profile · See the Johannesburg profile

Evidence: Descriptive / self-reported Top 43% 53/100 · Ask Evidence Copilot about this practice

The University of Johannesburg upgraded its MoUJi chatbot with LLM technology so students can check admission scores, register and get financial-aid help in isiZulu, Afrikaans, Sesotho or English; an earlier version cut peak-season staffing from 120 to 60.

100+
Common student-query types handled
120
Peak-season contingent staff, before
60
Peak-season contingent staff, after
R800,000+
Reported cost savings
4
Languages supported
MoUJi — University of Johannesburg's Multilingual AI Chatbot for Admissions, Registration and Financial Aid

Details

Maturity
Established
Promoter
University of Johannesburg
Period
2019–present (multilingual LLM upgrade launched August 2026)
Keywords
higher education administration, chatbot, student services, multilingual AI

Context

MoUJi is the University of Johannesburg's AI-powered virtual assistant, first launched in 2019 and substantially upgraded in August 2026 with large language model technology that lets it hold natural, context-aware conversations instead of relying on scripted, keyword-matched replies.

Objectives

The upgraded chatbot aims to serve prospective and current students in isiZulu, Afrikaans, Sesotho and English, helping them calculate their Admission Point Score, check admission requirements, apply, register for classes, and get information on financial aid and accommodation without navigating multiple webpages or waiting for office hours.

Activities

Institutional answers are pulled directly from UJ's official systems, and identity-sensitive queries use one-time-password verification. Before this LLM upgrade, an earlier version of MoUJi built on Amazon Lex already handled over 100 common student-query types.

Results

The earlier Amazon Lex-based version let the university reduce its peak-season contingent workforce from 120 to 60 staff, saving more than R800,000, according to an AWS case study. University leadership frames the tool as reducing pressure on the contact centre so staff can focus on complex cases.

Conclusions

This is a plausible, sourced operational-efficiency claim, though no independent third-party evaluation of learning or student-outcome impact exists, since MoUJi is fundamentally an administrative-support tool rather than a teaching one.

Implementation
Indicative cost
Medium (€50k–€500k)
Time to results
Medium (1–3 years)
Staffing & skills
University of Johannesburg IT and registrar's office team, AWS technical partnership

Conditions for success

  • Answers pulled directly from UJ's official systems for accuracy
  • One-time-password verification for identity-sensitive queries
  • Multilingual delivery (isiZulu, Afrikaans, Sesotho, English) matching the student population

Common failure modes

  • Reported cost/staffing savings come from a vendor (AWS) case study, not an independent audit
  • As an administrative tool, it has no learning- or teaching-outcome evidence

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful