# EdTech companies (/docs/use-cases/edtech-companies) # Moderate learner content without replacing educator judgment [#moderate-learner-content-without-replacing-educator-judgment] Evaluate discussions, classroom messages, learner profiles, and submissions against age- and course-appropriate rules while keeping educators and safeguarding teams responsible for contextual decisions. Learning products combine community interaction with users who may need different safeguards by age, course, or institution. Insion lets you express those boundaries through rules, retain classroom context, and connect content decisions to the right response in your product. ## Keep routine screening separate from safeguarding decisions [#keep-routine-screening-separate-from-safeguarding-decisions] Harassment can derail a class, spam distracts instructors, and exposure of private details creates escalation work. A universal blocklist cannot handle quoted source material, health education, or instructor-led discussion. Manual review of every submission becomes too slow and expensive as courses grow. With Insion, routine evaluation, durable records, user state, webhooks, and appeals share one integration. Educators and safety teams retain their time for decisions that need real educational context. ## Preserve course, classroom, and role context [#preserve-course-classroom-and-role-context] Moderate course discussions, peer replies, classroom messages, tutor conversations, profiles, usernames, avatars, written submissions, image attachments, and resource links. Keep the course, classroom, role, and age-band context attached to each decision. Check public contributions before they appear. Evaluate teacher-visible submissions in the background when the learning flow should continue without waiting. ## Allow curriculum content while identifying targeted harm [#allow-curriculum-content-while-identifying-targeted-harm] Hate and harassment, Privacy and doxxing, Self-harm, Violence, and Sexual content involving minors are relevant presets. Custom rules can allow curriculum-required discussion of difficult historical or health topics while flagging targeted abuse, attempts to expose another learner's private details, and sexual exploitation content involving anyone under 18. ## Use the result without replacing educator judgment [#use-the-result-without-replacing-educator-judgment] | Situation | Possible learning-product response | | -------------------------------------------------------- | ------------------------------------------------------ | | A public discussion post contains targeted abuse | Hold it before classmates see it | | A submission discusses difficult curriculum material | Preserve it for the instructor with course context | | Content raises a safeguarding concern | Route it through the institution's established process | | A specially handled learner or staff account is involved | Keep account action under human control | A moderation result is an input to the learning product, not a complete safeguarding decision. Educators and safeguarding teams remain responsible for cases that require knowledge of the learner, course, or institution. ## Keep automation connected to educator judgment [#keep-automation-connected-to-educator-judgment] A flagged result can hide a public post, preserve a submission for an instructor, or update your safeguarding process. Account-status changes can be applied consistently across classrooms. Your product remains responsible for enforcing that status. Mark staff or specially handled learners as protected from automated action. If appeals are enabled, an eligible suspended user can request review and your team can approve or reject it with the record context available. One moderation foundation supports different learning formats while custom rules preserve classroom context. Operational effort can stay focused on meaningful intervention rather than routine screening. ## Roll out with educators and safeguarding teams [#roll-out-with-educators-and-safeguarding-teams] Start with a focused area such as public course discussions, not every learner interaction at once. Test mode evaluates content and stores records while preventing changed results from sending moderation webhooks or automatically changing user status. Educators can review representative outcomes, verify that curriculum context is treated correctly, and refine custom rules before any learner-facing action occurs. Define separately what happens to a flagged post, the linked account, and cases that enter an existing safeguarding process. Protect teacher, administrator, and specially handled learner accounts when automatic suspension would conflict with that process. Avoid sharing sensitive student data unless it is necessary for moderation and appropriate under your own data practices. When the first area is stable, extend the same approach to classroom messages or submissions while preserving course and role context. Communicate the moderation boundary in language learners and educators can understand. A content status is an input to your learning product, not a complete safeguarding judgment. Preserve instructor access where your procedures require it, and route urgent or contextual concerns through the people and systems already responsible for learner welfare. This keeps automation useful without hiding institutional accountability behind a software decision.