From Custom Chatbots to a Scalable Instructional Platform for Language Teaching and Learning

A core challenge across language education is that students lack adequate opportunities for sustained, low-stakes conversational practice outside the classroom, a limitation that is particularly acute in fully online, asynchronous courses. This project addresses this conversational gap by designing, implementing, and assessing customizable conversational bots for French, Spanish, and German courses. Moving in two overlapping phases, Phase 1 will leverage an existing GenAI platform to build and test task-based bots that provide scaffolded practice, formative feedback, and role-play activities. Phase 2 will transition these empirical insights into developing an instructor-facing platform prototype, allowing faculty members to easily create course-specific bots using validated templates and configurable voice settings.  

• The Problem Addressed: Standard AI tools often over-explain or simply write plausible text on behalf of the student, failing to align with specific Common European Framework of Reference for Languages (CEFR) levels, course objectives, or cultural nuances.  
• The GenAI Approach: Bots are explicitly structured around controlled communicative tasks (e.g., recounting an experience, requesting a favor) utilizing structured 10–15 question sequences.  
• Advanced Audio Integration: The project integrates cutting-edge Audio and Text-to-Speech (TTS) APIs, allowing students to adjust parameters like speech rate to safely build oral accuracy and communicative confidence.  

As the project advances through its developmental phases, the team will publish and distribute a suite of practical tools for the broader UWaterloo community. Deliverables will be shared on this project website and via broader University-wide communication channels. Stay tuned!