NextSchool uses AI to suggest independent schools that fit your family. This page explains in plain language what goes in, how the matching works, and how to get a human to review or override anything you disagree with. Quebec's Law 25 (s. 12.1) and the federal Personal Information Protection and Electronic Documents Act both give you the right to know how automated decisions about you are made — this page is that disclosure.
What goes in: your inputs
Recommendations are built from the preferences you share with our AI consultant during chat. We use:
- Location and commute— the area you live or want to be near, and how far you're willing to travel.
- Grade and child profile — age/grade, learning style, interests, and any learning needs you choose to share.
- Priorities and dealbreakers — what matters most (e.g. academic rigour, arts, athletics, faith, special needs support) and any hard requirements.
- Budget— tuition range you're comfortable with, including whether financial aid matters.
- Curriculum and pedagogy — preferences for IB, Montessori, Reggio, Waldorf, traditional, etc.
We never sell or share your inputs with schools or third parties for marketing. See our Privacy Policy.
How the matching works
The matching process happens in three stages, each handled by a different backend pipeline:
- Filter. We narrow our database (independent schools across Canada) to schools that satisfy your hard requirements: location, grades offered, co-ed/single-sex, curriculum type, religious affiliation, etc.
- Score. Each remaining school is scored against the priorities you shared. Scoring blends:
- Structured signals (does the school explicitly support what you asked for? e.g. dyslexia program, French immersion, AP courses).
- Semantic similarity between your priorities and the school's self-described program, calculated with text embeddings.
- A second-pass cross-encoder rerank to push the most relevant matches to the top.
- Explain.A language model generates a short rationale for each top match — the trade-offs, the strongest fit areas, and what to watch out for. These rationales are guardrailed: we don't generate facts the database doesn't support.
What the system doesn't do
- It does not make admissions decisions. Schools decide who to admit.
- It does not score your child or your family. The score is on the school, against your stated priorities.
- It is not a ranking service. Two families with different priorities will see different orderings of the same schools.
- It does not track you across the web. We use first-party data plus Google's Consent-Mode-compliant ad measurement, which you can opt out of in Settings → Privacy.
Get a human to review
You can request a human review of any recommendation, scoring decision, or piece of information about a school at any time. Email [email protected] with the school name and what you'd like reviewed. We respond within 30 days (typically faster). You can also dispute facts about a school directly via the Report inaccurate information link on each school profile.
Your rights under PIPEDA + Law 25
- Access the personal information we hold about you.
- Correct inaccuracies.
- Request deletion (right to be forgotten).
- Export a copy of your data (data portability).
- Withdraw any consent you previously granted.
- Request a human review of an automated decision (Law 25 s. 12.1).
All requests go to [email protected] or the controls in Settings → Privacy.
Our Privacy Officer is the CEO of NextSchool. Mailing address available on request.
