The HALC AI Guide is designed to grow with its tutors and students. We track progress in three connected areas: tutor development, student learning, and overall program quality. Our approach is simple and built into the tools already in use. The WordPress microlearning section, the HALC onboarding modules, and the regular session reports already capture the data. Brief checklists, reflections, and observations show how AI is supporting learning in real time.
Tutor Development
Tutor growth is at the heart of the project. The goal is to help tutors use AI thoughtfully and effectively in their daily work. In Phase 1: Essential AI Skills Microlearning, hosted on the EdTech WordPress site, short quizzes and badges show tutors’ grasp of key concepts such as privacy, academic integrity, and ethical prompting. As they move to the HALC site for the Phase 2: Everyday Tutoring Skills, they begin reflecting on live sessions and completing short knowledge-check quizzes, describing how they used Socratic questions, CRA scaffolds, or feedback prompts. The session reports already implemented in HALC now include an AI usage section, seamlessly integrated into the existing workflow for easy tracking. Periodic scheduled observations give coordinators a fuller picture of how Copilot and custom apps are being used in sessions.
Across these stages, progress is measured through completed badges, reflection quality, and the visible use of questioning and scaffolding strategies during sessions. The emphasis is always on growth and self-awareness rather than evaluation.
Objective 1: Improve Student Learning Through AI-Enhanced Tutoring AI integration should deepen understanding, not replace it. Students will demonstrate stronger reasoning and problem-solving skills after working with AI-supported tutors or bots. KPI: Students show clearer understanding through improved quiz or concept-check results after tutoring sessions. Implementation / Evidence: Brief concept checks, Copilot-generated quiz questions, and recorded outcomes from existing HALC reports.
Objective 2: Model Responsible AI Use for Students Tutors will help students see AI as a thinking partner—one that supports exploration and reflection rather than giving answers. KPI: Student reflections and tutor observations indicate that AI was used to guide reasoning, not to supply solutions. Implementation / Evidence: AI usage notes in HALC session reports, tutor reflections, and feedback discussions during or after sessions.
Objective 3: Increase Student Engagement and Confidence Students will engage more actively in the learning process, ask deeper questions, and feel more confident applying what they’ve learned. KPI: Students report higher confidence and engagement in post-session feedback forms. Implementation / Evidence: Short post-session surveys or existing HALC feedback prompts measuring confidence and perceived clarity.
Objective 4: Encourage Continued Learning Beyond the Session Students will use AI responsibly and independently as a learning tool outside of tutoring sessions. KPI: Returning student rates and use of HALC bots or apps between sessions demonstrate extended engagement. Implementation / Evidence: Attendance records, follow-up surveys, and analytics from HALC custom bots or apps hosted on GitHub Pages.
Purpose: Capture how AI-supported tutoring occurs during the session. Implementation: AI usage field in the session report.
AI was used as a guide, not an answer source.
One or more strategies applied: ☐ Socratic questioning ☐ CRA scaffolding ☐ Feedback prompts.
Student engaged through questioning or reasoning before AI input.
Responsible use of AI modeled (tool limits or proper attribution discussed).
Opportunities for improvement or next steps noted.
Specific HALC Bot or App used (specify)
Purpose: Record brief observations of AI integration and engagement. Evidence: Observation logs or coordinator notes.
Tutor prompted AI with questions and verified responses.
Student interaction was active and reasoning-based.
AI identified as a support tool, not a replacement.
Session met privacy and academic-integrity expectations.
Summary / Notes:
Purpose: Guide creation process and evaluate tutor-built bots for quality and ethics. Evidence: Bot update notes or semester audit summary.
Bot guides learning instead of giving direct answers.
AI integration is valuable only if it helps students learn better and models responsible AI use for them. The custom bots developed in HALC guide learning rather than provide direct answers, emphasizing reasoning, exploration, and reflection. Students are encouraged to view AI as a supportive tool—one that extends learning opportunities inside and outside of tutoring sessions. To understand that impact, HALC uses light, built‑in feedback while showing students how AI can support, not replace, human learning. Short concept checks at the end of a tutoring session or quick forms in the HALC system capture what students understood, how confident they feel, and whether they plan to return.
The measure of success here involves a clearer understanding shown through improved quiz or concept check results, stronger engagement reflected in more active participation and detailed reflections, continued responsible AI use, and students who return for support. These outcomes are tracked through quiz data, session reflections, and attendance information already collected in HALC, avoiding extra reporting steps.
Objective 1: Model Responsible AI Use for Students Tutors will demonstrate and reinforce how AI can be used as a learning partner rather than an answer provider. KPI: Session reflections and observations show tutors using AI to guide student reasoning and discussion, not to generate final answers. Implementation / Evidence: AI usage notes within HALC session reports, brief reflection summaries, and coordinator observations of live or recorded sessions.
Objective 2: Encourage Active Participation and Reflection Students will engage with AI tools through questioning, reasoning, and self-explanation rather than passive use. KPI: Student feedback and reflection forms indicate higher engagement and awareness of their own learning process. Implementation / Evidence: Post-session check-ins or concept-check forms already embedded in HALC’s reporting system, combined with tutor reflections describing student interaction quality.
Objective 3: Support Independent Learning Beyond Tutoring Sessions Students will apply AI tools appropriately and confidently outside tutoring sessions, using HALC bots or apps to review or explore new problems. KPI: Returning student activity, continued use of HALC bots and apps, and responsible self-guided engagement reported through analytics or follow-up feedback. Implementation / Evidence: Attendance and usage data from HALC logs, GitHub analytics for custom bots or apps, and short follow-up surveys.
Student Learning and Engagement in AI-enhanced sessions are measured through aggregated, qualitative evidence. Individual usage data from CUNY Copilot and GitHub apps is not accessible, though system-wide Copilot activity reports may eventually be available from the CUNY Central Office.
HALC relies on Tutor AI-Enhanced Session Reports, coordinator summaries, and brief student feedback to understand how students reason, participate, and grow in confidence when using AI.
Purpose: Gather quick, anonymous feedback about how AI tools supported learning.
Implementation: To be embedded into the HALC site or emailed as a short Microsoft Form.
The tutor helped me use AI to think through problems, not just find answers. ☐ Yes ☐ Somewhat ☐ No
After this session, I feel more confident explaining how I reached an answer. ☐ Yes ☐ Somewhat ☐ No
I understand better how to use AI tools responsibly and ethically for learning. ☐ Yes ☐ Somewhat ☐ No
(Optional) What part of this session helped you learn the most? _____________________________
(Optional) What could make future sessions even better? _____________________________
Purpose: Gather quick, anonymous feedback about how AI tools supported learning.
Implementation: Aggregated results summarized in the coordinator’s semester notes and cited in the Program Quality Review Checklist.
Frequency: Once per semester (week 2–3 baseline; last 2 weeks follow-up).
Delivery: Microsoft/Google Forms link via email or QR in centers.
Have you used AI tools for learning (Copilot, ChatGPT, etc.) in the past 6 months?
Yes
No
Not sure
How comfortable are you using AI to explore ideas or check your understanding?
Very comfortable
Somewhat comfortable
Not comfortable
In a tutoring context, what do you think AI should mostly be used for?
Guiding thinking and problem-solving
Explaining concepts in different ways (examples, visuals, analogies)
Giving answers quickly
Not sure yet
Right now, how confident do you feel explaining your reasoning in STEM courses?
Very confident
Somewhat confident
Not confident
What’s one thing you hope AI can help you with this semester? (Open-ended)
Implementation: Aggregated results summarized in the coordinator’s semester notes and cited in the Program Quality Review Checklist.
Frequency: Once per semester (last 2 weeks follow-up).
Delivery: Microsoft/Google Forms link via email or QR in centers.
In HALC tutoring this semester, AI was used primarily to:
Guide my thinking and problem-solving
Explain concepts (examples, visuals, analogies)
Give answers directly
I didn’t notice AI being used
Compared to the start of the semester, my confidence in explaining my reasoning is:
Higher
About the same
Lower
I understand how to use AI responsibly for learning (not as an answer machine):
Strongly agree
Agree
Disagree
Outside of sessions, I used HALC bots or apps to study or practice:
Often
Sometimes
Rarely
Never
What part of AI-supported tutoring helped you most this semester? (Open-ended)
Program Quality and Ethics, and Responsible AI Use
Because the HALC AI Guide is supported by the ADELANTE and IDEAS grants, it also serves as a model for ethical, inclusive, and sustainable AI use at Hostos. Each semester, EdTech and HALC teams review quiz outcomes, reflection data, and accessibility checks from the WordPress and HALC sites. They confirm that all bots and tutor-built apps meet CUNY privacy standards and remain accessible to every learner. These regular reviews keep the project aligned with CUNY policy, FERPA compliance, and the broader mission shared with COTE—to uphold quality, equity, and innovation across learning and teaching at Hostos.
Objective 1: Ensure Ethical and Inclusive AI Implementation The HALC AI Guide will serve as a model of responsible, equitable, and sustainable AI use across Hostos. KPI: Regular reviews confirm that all AI tools, bots, and tutor-built apps meet CUNY privacy, FERPA, and accessibility standards. Implementation / Evidence: Semester audits conducted jointly by EdTech and HALC teams, including privacy, accessibility, and ethical use checklists.
Objective 2: Maintain Alignment with Institutional and Grant Goals The HALC AI Guide will continue to reflect the values of the ADELANTE and IDEAS grants and align with the COTE framework’s focus on quality, equity, and innovation. KPI: Documentation and updates demonstrate progress toward grant outcomes and COTE-aligned practices. Implementation / Evidence: Semester review summaries, internal reports, and updates shared between EdTech, HALC, and Institutional Research.
Objective 3: Promote Continuous Improvement and Transparency Program data will be used not for evaluation, but for growth—informing future microlearning updates, app revisions, and professional development. KPI: Adjustments to modules, bots, or training materials are made each semester based on review findings and tutor feedback. Implementation / Evidence: Version updates on WordPress and HALC sites, GitHub revision history for apps, and semester debrief meeting notes.
Purpose: Ensure Ethical and Inclusive AI Implementation; Maintain Alignment with Institutional and Grant Goals; and Promote Continuous Improvement and Transparency.
Implementation: Internal reviews are conducted each semester and filed with audit documentation. Results are shared with EdTech and HALC teams for continuous improvement and compliance tracking.
No student data is stored or shared outside the secure CUNY environment.
Accessibility review completed (keyboard navigation, alt text, and color contrast).
Tone and instructional content are inclusive and student-centered.
AI outputs are reviewed for accuracy and bias at least once per semester.
All materials are published under approved CUNY or open-license formats.
Activities align with grant objectives (STEM equity, innovation, digital literacy).
Updates shared with EdTech, Institutional Research, and grant reporting teams.
COTE framework principles (UDL, accessibility, inclusivity) are applied where relevant.
Program outcomes are reviewed in semester-end leadership meetings.
Feedback from tutors and coordinators is collected and discussed each semester.
WordPress modules are updated annually based on reflection or quiz data.
Version history or changelog is documented and accessible.
Semester summary report created and archived.
A Continuous Cycle
The HALC AI Guide treats data as a feedback loop, not a scorecard. Tutors see their own progress and can use it to plan next steps, while coordinators look at overall trends to refine training materials. Each quiz, reflection, and session report contributes to a shared understanding of what works best in AI-enhanced tutoring.
Need help with the Commons?
Email us at [email protected] so we can respond to your questions and requests. Please email from your CUNY email address if possible. Or visit our help site for more information:
