Kwame 2.0: Human-in-the-Loop Generative AI Teaching Assistant for Large Scale Online Coding Education in Africa
arXiv:2603.29159v1 Announce Type: new 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.
View PDF HTML (experimental)
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.
Comments: 8 pages, Accepted at the 27th International Conference on Artificial Intelligence in Education (AIED 2026)
Subjects:
Computation and Language (cs.CL); Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
Cite as: arXiv:2603.29159 [cs.CL]
(or arXiv:2603.29159v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2603.29159
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: George Boateng [view email] [v1] Tue, 31 Mar 2026 02:19:28 UTC (619 KB)
Sign in to highlight and annotate this article
Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
announceassistantvaluation
Proof of Usefulness Weight Distribution
The weights in the Proof of Usefulness algorithm were not assigned by committee consensus, philosophical preference, or gut feeling. They reflect the consistent findings of startup failure research — particularly the question of which early signals most reliably predict whether a project creates lasting value. This is what that research shows. Read All
Detecting Toxic Language: Ontology and BERT-based Approaches for Bulgarian Text
arXiv:2604.01745v1 Announce Type: new Abstract: Toxic content detection in online communication remains a significant challenge, with current solutions often inadvertently blocking valuable information, including medical terms and text related to minority groups. This paper presents a more nu-anced approach to identifying toxicity in Bulgarian text while preserving access to essential information. The research explores two distinct methodologies for detecting toxic content. The developed methodologies have po-tential applications across diverse online platforms and content moderation systems. First, we propose an ontology that models the potentially toxic words in Bulgarian language. Then, we compose a dataset that comprises 4,384 manually anno-tated sentences from Bulgarian online forums
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Research Papers
![Thoughts on AI and Research [pdf]](https://d2xsxph8kpxj0f.cloudfront.net/310419663032563854/konzwo8nGf8Z4uZsMefwMr/default-img-robot-hand-JvPW6jsLFTCtkgtb97Kys5.webp)
Request for arXiv cs.AI Endorsement – Life-Aligned AI Framework
Hi everyone, I’m preparing to submit a paper to arXiv (cs.AI, with cross-lists to q-bio.PE and physics.soc-ph) and am currently awaiting endorsement from a qualified author. Posting here in case anyone in this community can help or knows someone who can. Title: Life-Aligned AI: A Framework for Grounding Artificial Intelligence in the Empirical Conditions of Flourishing Here’s the main idea: Current alignment approaches work backwards — rules imposed in advance by minds that the systems they constrain may eventually exceed. This paper proposes a different starting point: training AI on living systems — the only adaptive framework continuously pressure-tested across four billion years under conditions of genuine consequence — and letting the operating principles emerge from genuine reasoning
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!