Open-source modular AI math tutoring platform

Product or Programme · All products and programmes
Evidence · 2 claims (1 for, 1 mixed) · 1 study (1 causal), q2 · 0 of 1 report an effect size · 2 claims rest on one study

Description

An open-source modular platform developed by Jarosław A. Chudziak and colleagues for AI-powered personalized mathematics tutoring, integrating conversational tutoring, retrieval, task generation, symbolic solving, plotting, and course-graph tools.

Components

• AITutor: an LLM-based iPad tutoring app translating pedagogical mechanisms into UI features: AITutor is a React Native/Expo iPad app with a Python/FastAPI backend that orchestrates LLM calls, curriculum-constrained prompting, retrieval over digitized course materials, and telemetry logging. Rather than "treating the AI as an open-ended chatbot, we constrained the interaction, translating pedagogical mechanisms into explicit UI features." Its features are layered worked examples (overview, orientation, explanation, final-answer checkpoint, exam-format layer), step-linked visual grounding, a contextual follow-up dock, and a delayed retrieval/wrong-book loop. (From Answer Generators to Reasoning Facilitators: Designing AI Tutors for Mathematical Rea)
• MathBuddy: multimodal emotion-aware LLM math tutoring system: MathBuddy is an emotionally aware LLM-powered math tutor that, per the authors, "dynamically models the student's emotions and maps them to relevant pedagogical strategies, making the tutor-student conversation a more empathetic one." It captures emotions from conversational text (a fine-tuned DistilBERT classifier) and from webcam facial expressions (face-api.js, reporting 76% accuracy and 71% F1 on its test set), aggregates both modalities into Positive/Neutral/Negative, and prompts the LLM tutor accordingly. The system has a TypeScript frontend with chat interface, webcam input, and visual workspace, and a Python REST backend for inference. Code, dataset, and demo video are publicly released. (Debanjana Kar et al. (2025))

Claims

• In a within-subject user study, participants showed longer positive facial expressions under the emotion-aware (Emotion ON) tutoring condition [+M]
• MathBuddy's multimodal emotion aggregation matched participant-reviewed emotions at 60% overall accuracy, with sharply reduced recall for the Neutral class [~M]

Related Products and Programmes

• Math Mysteries

Key Sources

• Jarosław A. Chudziak and Adam Kostka. (2025). AI-Powered Math Tutoring: Platform for Personalized and Adaptive Education. arXiv preprint. https://arxiv.org/abs/2507.12484
• Harry Feng, Yuan Tian, and Erica Zhao. 2026. From Answer Generators to Reasoning Facilitators: Designing AI Tutors for Mathematical Reasoning in High-Stakes Environments. https://arxiv.org/abs/2607.01692
• Debanjana Kar, Leopold Böss, Dacia Braca, Sebastian Dennerlein, Nina Hubig, Philipp Wintersberger, and Yufang Hou. (2025). MathBuddy: A Multimodal System for Affective Math Tutoring. https://arxiv.org/abs/2508.19993

Page metadata

Field	Value
type	product
id	open-source-modular-ai-math-tutoring-platform
title	Open-source modular AI math tutoring platform
description	An open-source modular platform developed by Jarosław A.
status	review
product_kind	software
generated	by process:settle-candidates on 2026-10-10
trust	unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.