Speakers
Short summary
Personalizing language instruction is challenging. This session introduces a model for creating custom AI tutors that provide individualized support. Instructors modify NCSSFL-ACTFL Can-Do Statements to serve as writing prompts for activities. The AI provides instant feedback on grammar and syntax, eliminating student waiting periods. This allows educators to focus on high-level mentoring. You will learn to build these tools using Microsoft 365 without requiring a technical programming background.
Abstract
Providing timely, individualized support to students at varying proficiency levels remains a primary pedagogical challenge in language instruction. This presentation introduces a versatile model for developing custom AI agents that serve as “on-demand” tutors, ensuring every student receives personalized instructional support regardless of their initial proficiency. Moving beyond generalized instruction, this model uses NCSSFL-ACTFL Can-Do Statements as instructor-led writing prompts, allowing students to independently verify the accuracy of their responses through immediate AI-driven feedback. Attendees will move beyond the theoretical to a practical, step-by-step demonstration of “pedagogical engineering.” The session will demonstrate how to transform proficiency descriptors into writing activities and how to customize AI agents to provide students with instant feedback on grammar, syntax, and register. Crucially, the session highlights how these tools mitigate the "waiting period" that often stalls progress, replacing it with a safe, iterative environment where errors become opportunities for immediate refinement. By the end of the session, participants will understand the technical process of building these AI tutors using the Microsoft 365 ecosystem, demonstrating that effective customization is accessible to educators without requiring a technical programming background. This approach empowers teachers to reclaim their time for high-level mentoring while the AI manages the high-frequency aspects of formative feedback. Participants will leave with a clear blueprint for integrating these agents into their existing curricula to provide a comprehensive and supportive writing aid for their learners.
Keywords
AI feedback, Can-do descriptors, Writing instruction
Special scheduling requests
We have two co-presentations and a back-to-back schedule would be greatly appreciated.
| Scheduling preference | Anytime on Saturday |
|---|---|
| Title | Integrating Personalized AI Agents into the Writing Curriculum |