Make Learning Copilot work for your institution.

Use your normal Edumatica sign-in and institution/role selection. Learning Copilot setup belongs inside the product, not in the public website and not in the platform BackOffice OTP screen.

1. Sign in at app.edumatica.io/login, or your approved institution microsite. Complete the normal login with your own authorized account.
2. Select the institution owner or organization administrator role for the intended institution. Confirm the selected institution before changing anything.
3. Open Learning Copilot from the account menu. If the menu is not yet visible, open /learning-copilot on that same product or microsite host after signing in.
4. Check that the institution, staff and learner records are active. Faculty must be assigned to the intended learners through the current section/class or course assignment; a same-institution login alone is not enough.

This is the account-level switch that makes the reviewed practice workspace available. It does not automatically authorize live AI, buy tokens or change academic or financial records.

1. On the first-time card, select Enable approved practice. If practice is already enabled, continue to the next step; do not disable and re-enable it as a routine fix.
2. Agree the institution learning-day time zone before the first rollout. The current first-time web action uses the server-provided zone (Asia/Kolkata by default); it does not display a time-zone picker. Ask support to review a different required zone before enabling. Once configured, the zone is locked.
3. Start with a small, clearly assigned cohort and one subject. Keep the existing courses, attendance, assignments, marks and fees unchanged.
4. If practice must be paused later, an authorized owner/admin can use Settings → Pause approved practice. Pausing is not deletion and is not a way to reset learning history or spend.

The Library accepts a short plain-text passage. Publishing a reference is a deliberate staff action. A saved draft is not yet a learner-visible source.

1. Open Library → Add reference material.
2. Enter Source title, Source subject and Reference passage. Use material you own or may use for these learners; this workflow does not extract PDFs or scan files. The passage limit is 12,000 characters.
3. Select the Authorised learners. Use Load more source learners if the intended learners are on a later page. Confirm the rights checkbox: I own this material or have permission to use it for these learners.
4. Select Save source for review. Read the saved passage, confirm the audience and content, then select I reviewed this source — publish.

Choose a small set of objective questions your staff can verify. For a first rollout, three to five questions are usually easier to review than a large bank.

1. Open Practice packs → Create practice pack. Enter Pack title, Subject and Learning objective.
2. If the ERP section/course/academic-period setup exists, select Choose ERP learning context and the correct ERP learning context. Every selected learner must belong to that context. Otherwise keep General practice (unmapped), with that limitation clearly understood; do not invent a course or year.
3. Under Assign learners, use Find a learner and select only the intended audience. Faculty see their authorized assigned learners; use the available paging control when needed.
4. Select suitable published references if you are grounding the pack in material. A manual objective pack can be created without a reference; adult AI needs approved references.
5. For each item enter Question, two to four answer options, Correct answer, Hint (guide, do not give the answer) and Explanation shown after success. Packs support one to twenty questions.
6. Select Save draft for review. Confirm that the saved draft has the intended title, audience and learning context.

Publishing is the hand-off from staff preparation to learner practice. An AI-generated draft, if used, goes through this same human review.

1. Open the draft. Read every prompt, answer option, correct answer, hint and explanation. Correct ambiguous questions and inappropriate material using Edit draft.
2. Confirm the selected learners, the context and the continued availability of any attached references.
3. Check I checked every question, answer key, hint and learner assignment, then select I reviewed this pack — publish.
4. Confirm the pack status is published. Published questions are immutable. Use Create a new version for a reviewed replacement; Withdraw pack removes current learner access while preserving history.

Use an approved training learner for acceptance. Never reset a customer password or submit work as a customer merely to demonstrate the feature.

1. Sign in separately as the assigned learner and open Learning Copilot → Practice. Open the newly published pack.
2. Make an honest attempt and select Check my answer. Confirm that the reviewed explanation appears after success. On a different item, use Get a hint and check that guidance is shown.
3. Move between questions, return to the pack list and reload. Confirm that completed practice persists and repeated submission does not award points again.
4. Open Today and Progress to review available practice/revision and independent versus assisted evidence. In Settings, try Show learning points and rewards and Reduce motion using the training account.

Staff evidence tells you how practice happened. It is not an automated assessment of a learner’s ability or a substitute for teaching judgment.

1. Return to the published pack as its authorized author or manager and select Review learner progress.
2. Read completed items and XP in context. Use Review topic evidence for the intended learner to distinguish independent and guided practice.
3. An owner/admin may configure Settings → Learning revision calendar and Save learning calendar. Learners can adjust their own available learning preferences and revision controls.
4. Use rest days and calendar pauses deliberately. They affect the learning revision schedule, not ERP attendance, class timetables, fees or promotion.

Parent access is a reviewed snapshot, not an unrestricted feed of attempts or AI conversations. Current account relationships are checked at read time.

1. Confirm the parent is actively linked to the correct learner through the normal ERP parent relationship.
2. For a College learner, first ask the learner to choose Settings → Allow my linked parent to see teacher-published practice summaries. A parent record alone is not consent.
3. As authorized staff, open Review learner progress on the published pack, then select Publish parent summary for the intended learner.
4. Sign in as the approved linked parent training account. Open Learning Copilot, choose the linked child if prompted, and select View teacher-published progress.
5. Check that only the explicitly shared summary is visible. Private attempts, unpublished drafts, staff keys and private AI output must not appear.

This step is optional. Reviewed practice already works without it. Live help requires a qualified server provider, an approved institution policy, finite cost allowance and current role eligibility.

1. As owner/admin, open Settings → AI policy and spending. Read Used, Held for reconciliation and Remaining. Budget values are service costs in USD—not student fees or subscription charges.
2. Enter an approved Monthly institution cost ceiling (USD), Daily per-user cost ceiling (USD) and Policy approval reference. Use an internal approval reference, never a provider key or personal document.
3. Where your approval is valid, select Request live AI for eligible adults. Approve Faculty and administrator users on this route are verified adults with approved use only after that statement is true. Select Save AI policy.
4. For an adult learner, choose Select learner for AI eligibility. Confirm adulthood through your approved process, record current consent and an Eligibility verification reference, then Save learner eligibility. College enrolment is not proof of adulthood; do not approve minors to unblock a demo.
5. As an eligible adult, open AI preparation (staff) or Guided help (learner). Select up to four approved references, enter the question/objective and what you have tried, choose English or Hindi, then select Get guided help.
6. Staff may choose Editable practice draft as the Preparation task and select the intended learners. Generate private practice draft, inspect it and Save as an editable draft. Correct its content and complete steps 4–5 before learner publication.

Support should preserve the account boundary and the saved request, rather than turn a network error into duplicate content or another paid generation.

1. On a pack, reference or available AI result, a learner/staff user can select Report a learning concern, describe the issue and approve Share this report with my institution’s owner/admin. Select Send learning report.
2. The current queue reaches the institution owner/admin, not an automatic developer inbox. Managers review Learning reports in Settings, document the response, check I reviewed the report and this response, and use Resolve reviewed report.
3. If AI shows an earlier unresolved paid request, select Check saved request. Do not create a new request, switch accounts, clear browser state or retry generation repeatedly to bypass pending/uncertain spend.
4. After an ambiguous manual source/pack create or copy, inspect the owned list before repeating the write. If its outcome is still unclear, stop and escalate with a timestamp and the resource/request identifier.
5. Record the institution, rollout scope, responsible staff, pack/source references, acceptance results and support owner in your secure internal handover. Do not put passwords, provider keys or private prompts in this public guide or a shared ticket.