How the question box works — and how it protects students

When you ask a question here, the answer comes from a library of 145 published best-practice guides written for paraeducators — not from an AI making things up. This page explains what happens to your question, what the tool will and won't do, and how students are protected.

Your question stays private unless you say otherwise

• Asking is private by default. Your question is not posted anywhere public.
• If you choose to share it to the public board, the public version is built from fixed dropdown choices (your role, the general type of support context) and a short general description — the form is designed so student-identifying details don't fit.
• Before anything is shown publicly, an automatic pass also softens names to initials and removes email addresses and phone numbers — every time, on our servers.
• Small-group protection: if fewer than 10 people share a combination like role + county + disability context, we don't display that combination — the county becomes "Washington" and the disability context is hidden. It's the same kind of protection researchers use to keep survey answers anonymous.
• The public board is excluded from search engines.

Safety questions get the safety path

• Questions about seizures, choking, severe allergic reactions, poisoning, suicide or self-harm, abuse, restraint, medication, or IEP decisions are routed to a safety response — not a casual answer.
• Crisis resources (the 988 Suicide & Crisis Lifeline, the Crisis Text Line) and call-911 guidance appear on those answers no matter what — they don't depend on the matching being perfect.
• Where an answer includes a hard safety threshold (like when a seizure becomes a 911 call), the words come verbatim from the underlying guide and are checked automatically every time the site is rebuilt.
• Every answer respects your scope of practice: what you can do right now, and who owns the decision — the nurse, the BCBA, the case manager, your admin. This tool will never coach you to diagnose, give medication advice, or make IEP calls.
• Clinical first-response steps appear only after an independent reviewer (like a registered nurse) signs off on that content. Those reviews are in progress; until each one lands, you'll see safe routing and "bring this to your team" instead.

What AI does here (and what it doesn't)

• Answers are not generated by a chatbot. The tool matches your question to the library and shows you what the library says — the same question gets the same answer, and a person can review every answer it's possible to get.
• The one place AI runs: if you choose to publish your question, an AI may suggest a more general rewrite with identifying details removed. You approve, edit, or reject it — and the automatic protections above run regardless of what the AI said.
• Your questions are not sold and are not used by Peerlore to train AI models. No account is needed. Nothing here asks for student records or names.

Why not just ask a chatbot?

A general AI chatbot doesn't know what a paraeducator is and isn't allowed to do, will improvise medical or legal answers, won't reliably treat a seizure question as an emergency, and invites you to paste a student's story into a third-party service. This tool is built the opposite way: a fixed library experts can review, safety routing that doesn't improvise, and pressure at every step to keep students unidentifiable.