Getting Started with the AI-PLC Agent
Keep this guide open as you move through the application and review each stage of the workflow.
Getting Started
Learn how to put the AI-PLC Agent to work.
Practical guidance and videos for preparing authorized inputs, choosing a protocol, using the PLC Assistant to question and refine the work, and reviewing the completed report before planning your next instructional response.
Before You Begin
Before You Begin
Confirm that your school or district permits the tool, the protocol, and the information you plan to provide. Start practice sessions with synthetic examples that do not represent actual students. A teacher-only interface does not override local restrictions on AI assessment or student-data use.
Featured guide
Getting Started with the AI-PLC Agent
Follow the AI-PLC Agent workflow with step-by-step guidance for creating rosters, adding Tasks and Success Criteria, uploading student work, checking inputs, using the PLC Assistant, and planning instructional responses.
How-To Videos
Find the workflow you need.
Practical guidance for setting up the app, building instructional clarity, working as a PLC, and getting unstuck when an analysis pauses.
Getting Started
Start with the practical setup steps you will use across the class-analysis protocols.
How to Create and Edit Class Rosters
Create and maintain the roster used to match work to the correct learner. Include only authorized information needed for matching, and check names and submissions before analysis.
How to Upload and Analyze Student Work
Provide the Task, Success Criteria, and authorized work samples, then select a Class Snapshot, Class Action Report, or Teacher Playbook. Check student matches before the run and review the generated analysis against the work afterward. The PLC Assistant can also be used during analysis to ask questions, request alternatives, and revise draft feedback while names remain masked.
Building Clarity
Use the Agent before instruction to develop a shared understanding of what students are expected to learn and demonstrate.
Getting Started with Unpacking for Clarity
Develop draft learning targets, Success Criteria, questions, and instructional ideas. Review them against the standards, curriculum, and learning experience before using them.
Working as a PLC
Bring class-level evidence together to identify patterns and decide what the team will do next.
Run the PLC Mastery Pulse Report
Bring class reports together for collaborative inquiry. Check the source evidence, Tasks, and Success Criteria before comparing patterns or choosing a shared instructional response.
Troubleshooting
Quick help when something does not move forward as expected.
What to Do When the Agent Stalls
Troubleshooting guidance for a run that pauses before completion. The PLC Assistant also supports everyday questions and revisions; see Use the PLC Assistant Throughout the Protocol below. When requesting support, use the run identifier and only authorized information. Avoid sending student work or named reports through an unapproved channel.
PLC practice resource
AI-PLC Simulation Protocol
Practice collaborative inquiry with a sample Class Action Report. Use discussion questions to examine evidence, consider instructional options, review draft feedback, and determine next steps together.
AI-PLC Simulation Protocol
A facilitated practice experience for examining evidence and planning an instructional response together.
See the outputs
See What the Reports Look Like
Explore example reports before running a protocol or reviewing your team's results.
View Example Reports (opens in a new tab)Work the Protocol
Use the PLC Assistant Throughout the Protocol
The PLC Assistant is the chat feature you use to work with the Agent's analysis and recommendations. It is not just for troubleshooting. Ask for more explanation or information, request more or less detail, explain why you disagree with a recommendation, or ask for another set of instructional strategies.
You can also revise draft Glows, Grows, and reflection questions while student names remain masked. Refer to the exact randomized student ID shown in the analysis when requesting a change to feedback for a particular response. These IDs do not correspond to roster or upload order. You do not need to know the student's name to ask for clearer, more specific, or better-supported feedback.
Keep requests connected to the original Task and its established Success Criteria. Do not add names or other identifying details to the chat to reconnect an ID with a learner. When discussing a band, describe the evidence on this task rather than a student's overall ability.
Ideas for What to Ask
| Purpose | Example Request |
|---|---|
| Ask for an explanation | What evidence supports this recommendation? Explain how it connects to Success Criterion 2. |
| Add practical detail | Expand this strategy with teacher questions, a brief example, and a short practice task. |
| Reduce detail | Condense these recommendations into three practical next steps without changing their meaning. |
| Disagree and request alternatives | I disagree with this approach. Suggest two different strategies that address the same Success Criterion and explain the difference between them. |
| Revise feedback while names are masked | For [displayed student ID], revise the Grow so it identifies one specific next step supported by the response and connected to the Success Criteria. |
| Adjust feedback wording | For [displayed student ID], keep the evidence-based meaning but use shorter sentences and a reflection question. Do not change the Success Criteria. |
These are example requests for educators to read and adapt. They are not connected to live application functionality and do not automatically transmit any requests.
Check the Completed Report
Once names are restored, compare student-specific suggestions with the original work and Success Criteria. Check that the revisions you requested are reflected accurately, and decide what feedback and instruction to use. Asking for a change does not remove the need to review the revised output.
If a Run Pauses Before Completion
You can also use the PLC Assistant to ask the protocol to continue its unfinished work. For example: Please continue the unfinished section using the Task and Success Criteria already provided. If it does not resume or the result is incomplete, follow the troubleshooting guidance. Do not assume a conversational request has repaired an application error.
Review Before Acting
Before You Use a Report
Check the original work, not just the summary. Treat performance bands as descriptions of task evidence, feedback as suggestions for review, and report groupings as planning possibilities. Educators determine what is communicated and what instruction occurs. Do not use a function that your district prohibits, even when its output will be reviewed.