Kwame 2.0: Human-in-the-Loop Generative AI Teaching Assistant for Large Scale Online Coding Education in Africa

arXiv cs.CL / 4/1/2026

📰 NewsSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces Kwame 2.0, a bilingual (English-French) generative AI teaching assistant for large-scale online coding education in Africa.
Kwame 2.0 uses retrieval-augmented generation to pull relevant course materials and generate context-aware answers within a human-in-the-loop forum in SuaCode.
In a 15-month longitudinal deployment covering 15 cohorts, 3,717 enrollments, and learners across 35 African countries, evaluations found the assistant delivered timely, high-quality curriculum support.
Human facilitators and peer participants helped reduce errors, with particularly strong mitigation for administrative queries where reliability demands are high.
The authors argue the human-in-the-loop design effectively combines AI scalability with human reliability to improve learning assistance in resource-constrained settings at scale.

Abstract

Providing timely and accurate learning support in large-scale online coding courses is challenging, particularly in resource-constrained contexts. We present Kwame 2.0, a bilingual (English-French) generative AI teaching assistant built using retrieval-augmented generation and deployed in a human-in-the-loop forum within SuaCode, an introductory mobile-based coding course for learners across Africa. Kwame 2.0 retrieves relevant course materials and generates context-aware responses while encouraging human oversight and community participation. We deployed the system in a 15-month longitudinal study spanning 15 cohorts with 3,717 enrollments across 35 African countries. Evaluation using community feedback and expert ratings shows that Kwame 2.0 provided high-quality and timely support, achieving high accuracy on curriculum-related questions, while human facilitators and peers effectively mitigated errors, particularly for administrative queries. Our findings demonstrate that human-in-the-loop generative AI systems can combine the scalability and speed of AI with the reliability of human support, offering an effective approach to learning assistance for underrepresented populations in resource-constrained settings at scale.