How Tori Teaches Every Student Differently — Special Education Accommodations in SetFlow

The problem with one-size-fits-all AI tutors

Pick any popular AI tutor today. Try it as a student with dyslexia. It will hand you long paragraphs, dense sentences, and idioms it never explains. Try it as a student with ADHD. Same wall of text. Try it as an English-language learner. Same idioms. Same assumptions.

A great teacher knows their students. A great AI tutor has to do the same — but it has to work the first time, before any “trust me, I've known her since September” has built up. The only way to get there is to put accommodation data in the model's hands.

What SetFlow's accommodation system does

Teachers set accommodations once on the class roster. Tori reads the accommodation profile on every interaction with that student — every Tori session, every quiz feedback, every rubric explanation.

SetFlow supports 16 accommodation types, with explicit teaching changes for each:

• Dyslexia — short sentences, bullet points, TTS auto-play on every Tori reply, no walls of text.
• ADHD — short energetic responses, micro-step task breakdown, break reminders.
• Autism Spectrum Disorder — literal language, no sarcasm or idioms, predictable structure, explicit step-by-step.
• Anxiety — no pressure language, effort acknowledged before content, mistakes normalised.
• Dyscalculia — visual number representation, tiny math steps, no time pressure.
• Dysgraphia — minimise required writing, accept verbal explanations.
• Executive Function disorder — explicit structure, checklists, previews of upcoming steps.
• Slow Processing Speed — one concept at a time, never rushed.
• Vision Impairment — TTS auto-play always on, visual content described in text.
• Hearing Impairment — text-first, visual instructions.
• Motor Disability — voice input prioritised, keyboard navigation.
• Auditory Processing — always text-based, visual preferred.
• ELL (English Language Learner) — simplified vocabulary, no idioms, terms defined inline.
• Gifted & Talented — harder follow-up questions, deeper exploration, faster pace.
• IEP / 504 — formal plan recognised; extended-time multipliers (1.5x, 2x, 3x) applied to every timed quiz automatically.

Two students with the same diagnosis get different Toris, because they are different people. Diagnosis sets the starting point; the adaptive learning model evolves from there based on what actually works for that specific student.

How a teacher sets accommodations

From any classroom roster:

1. Click the student's name.
2. Tick the accommodations that apply. Add a plan type (IEP, 504, ELL, Gifted, or Informal).
3. Optionally add private teacher notes for Tori (e.g., “extra patient when math word problems involve money”).
4. Save.

From that point forward — in this class, in every other class the student joins, in Study Mode, in Academy — Tori applies the accommodations. The student never sees the toggle and is never told their accommodations are active. The teaching just feels natural to them.

The adaptive learning model

On top of the accommodation profile, Tori builds a per-student learning model that updates after every session. The model tracks behavioural patterns — not conversation content. After roughly 25 sessions, it reaches mature confidence and Tori starts personalising in ways that no static accommodation profile can:

• Which explanation styles land for this student (analogy vs definition vs step-by-step).
• Their average pace and ideal session length.
• Their strongest and weakest concepts within the curriculum.
• What time of day they tend to learn best.
• What triggers frustration; what triggers breakthroughs.

The teacher can see this on the Student Intelligence Dashboard. The student can see a private, friendly version of it on their own dashboard after their fifth Tori session.

Privacy: where accommodation data is stored

Accommodation data is some of the most sensitive student data a school holds. SetFlow treats it that way:

• Under BYODB, accommodation rows live in your school's own database — not on SetFlow's servers.
• Tori uses the profile only to personalise its teaching. It is never shared with other institutions and never used for any commercial purpose.
• When the student account is deleted, the accommodation record is deleted with it.
• For under-13 accounts that signed up directly (outside a school deployment), the adaptive learning model is disabled by default — COPPA minor mode.

How this differs from Khanmigo and ChatGPT

Khanmigo and ChatGPT are general-purpose AI tutors. They don't know your school's accommodation framework, your student's IEP goals, or the specific lesson the student is working on. They can't — they live outside the LMS.

Tori lives inside the LMS. It reads the same accommodation table the teacher set up. It sees the same lesson body the student is reading. It builds a learning model that stays in your school's database for the student's entire enrolment. That depth of context is the difference between an AI that talks at a student and an AI that teaches this specific student.

How to get started

Schools: see /companies for the free 90-day institutional pilot. Teachers: try Tori free at sign up and explore the accommodation system on a test classroom.

More to read:

• Special Education in SetFlow — the full feature page.
• Adaptive Learning Intelligence — how the per-student learning model works.
• What is BYODB? — where accommodation data lives.