A Bilingual, LLM-Mediated Lecture Companion for Self-Regulated Learning: Architecture, Theoretical Framework, Comparative and Usability Evaluation, and a Pre-Registered Outcomes Protocol

Synthesis: Malhotra (2026) describes a browser-based lecture companion built for bilingual and connectivity-limited contexts, specifically the English-Hindi and 'Hinglish' mixing common in Indian classrooms. It records live lectures, transcribes them in the browser, and uses an LLM to turn transcripts into structured notes and study material, addressing conditions that most published systems ignore: tidy single-language English audio and reliable internet. The paper presents the architecture, theoretical framework, comparative and usability evaluations, and a pre-registered outcomes protocol. It advances Self Regulated Learning support and Accessible Learning for multilingual and resource-constrained Higher Ed settings, connecting to Language Learning and the Digital Divide.

Built for Real Classroom Conditions

Most published LLM lecture-note systems assume tidy, single-language English audio in well-connected settings — conditions that exclude many students. Across much of undergraduate education outside North America and Western Europe, instructors move between English and a home language mid-sentence, and reliable internet access cannot be assumed in classrooms or hostels. This lecture companion is designed for those conditions.

Architecture and Bilingual Transcription

The browser-based system records lectures in English, Hindi, or Hinglish (the English-Hindi mixing characteristic of Indian classrooms), transcribes them in the browser, and uses a language model to turn the transcript into structured notes and study material. On-device transcription addresses connectivity constraints, making the tool usable where cloud dependence would fail.

Supporting Self-Regulated Learning

The companion operationalises Self Regulated Learning by turning ephemeral lectures into persistent, structured learning resources that students can revisit and reorganise. Its comparative and usability evaluations, plus a pre-registered outcomes protocol, connect LLM study support to Language Learning, Accessible Learning, Accessible Learning, and equity goals in resource-constrained Higher Ed, speaking directly to the Digital Divide.

Connected Concepts

• LLM
• Self Regulated Learning
• Language Learning
• Accessible Learning
• Higher Ed
• Generative AI
• Digital Divide
• Student Experience
• Equity In AI Education

Connected Articles

• AI Guided Learning Audiovideo 2026
• Scheu Mobile Chatbot Journaling Motivation 2026
• Kutti AI Voice First Learning Companion
• International Students Conversational AI Adaptation
• Savvy Student Attention Video Learning
• GenAI Differentiated Eap Reading Materials 2026

Citation

Malhotra, T. (2026). A bilingual, LLM-mediated lecture companion for self-regulated learning: Architecture, theoretical framework, comparative and usability evaluation, and a pre-registered outcomes protocol. EdArXiv preprint.