The Neurodiverge AI Coach

Why we’re building it

Generic AI chatbots are unreliable for neurodivergent questions. They’ll happily quote ABA principles, recommend masking strategies, suggest BetterHelp, hedge every answer with “consult a professional”, and miss the lived-experience nuance entirely. The information is technically correct and emotionally useless.

What ND adults actually need:

• Answers grounded in ND-affirming material, not random internet content
• Identity-first language by default
• Honest, specific responses — not hedge / fluff / apology answers
• Recognition of the AuDHD and co-occurring patterns that get missed elsewhere
• Practical, actionable suggestions, not platitudes
• The ability to ask “why” and “what about” and have the coach actually engage

The Neurodiverge AI coach is built on top of the same 70+ guide library that powers this site. When you ask a question, the coach pulls from material written by ND adults, not generic web content. The answers are grounded, honest, and identity-aligned.

What you can ask

Anything that fits the lived experience of being neurodivergent. Examples:

• “I keep crashing every few months. Is this autistic burnout or depression?”
• “My partner thinks I’m being dramatic about lights and noise. How do I explain sensory load?”
• “I’m starting ADHD meds next week. What should I track to give my prescriber good feedback?”
• “I think I’m AuDHD but my therapist only sees the autism. What questions should I ask?”
• “I’m late-diagnosed autistic at 38. Where do I even start?”
• “My kid is autistic. The school wants ABA. What are the alternatives?”
• “I have an OT appointment tomorrow. What sensory features should I describe?”
• “Why does perimenopause feel like my masking suddenly stopped working?”
• “I got a BPD diagnosis years ago but I think it’s autism. How do I bring this up with my psychiatrist?”
• “I just took the AuDHD test and scored high on both. Now what?”

The coach will pull from the relevant guides, give a grounded answer, link the source material, and let you follow up.

How it works

• Retrieval-augmented. Every answer is generated from the relevant section of our guide library — not from open-internet AI training data. You can click through to the source material.
• Identity-first by design. The model is instructed to use identity-first language, ND-affirming framing, and to avoid pathologising or “fix yourself” defaults.
• No clinical jargon. Plain language, same voice as the guides.
• Honest about limits. When the question is genuinely outside the coach’s remit (crisis, specific medication dosing, formal diagnosis), it’ll say so and point at the right resource — without the reflexive “consult a professional” that turns up everywhere else.
• Conversational memory inside a session. You can ask follow-ups. “What about for women?” “What if I’m AuDHD?”
• No saved chat history by default. Sessions are ephemeral unless you opt into saving for later reference.

Who it’s for

• Adults newly recognising they’re neurodivergent and figuring out what it means
• Late-diagnosed adults rebuilding self-understanding
• AuDHD adults navigating a complex profile
• Parents of neurodivergent kids who need ND-affirming framing alongside the clinical material
• Anyone tired of generic AI giving generic answers to ND-specific questions
• Anyone preparing for a clinical appointment, school meeting, or work conversation about their neurology

What it’s not

• Not a diagnosis tool. The coach can’t and won’t diagnose. For that, you need a qualified clinician.
• Not a crisis service. If you’re in crisis, please contact a crisis line. The coach will recognise crisis signals and point you toward appropriate help, but it’s not a substitute for a human professional in acute moments.
• Not a replacement for therapy or community. A coach is one input among many. The best support is multi-channel.
• Not a medication prescriber. Medication conversations happen with prescribers. The coach can help you prepare for them.
• Not free of error. AI can be wrong. The coach is grounded in our guide library, which reduces error compared to generic AI, but isn’t infallible. Use judgment.

When it’s coming

• We’re building through 2026.
• Closed beta opens to a small group first. Soft launch after.
• We’ll announce on this page when the beta opens. Bookmark or check back.
• Until then, our self-screens and our growing guide library are free.

Pricing (planned)

• Free tier (planned). A small number of questions per day, no saved history. Lets anyone try the coach without a paywall.
• Paid tier (planned). Unlimited questions, saved chat history (opt-in), pattern continuity across sessions, integration with the upcoming tracker. Pricing will be confirmed when we launch.

We won’t run ads. We won’t sell data. The subscription pays for the work and the underlying AI compute.

Privacy

• Sessions are ephemeral by default. You opt in to history if you want it.
• We don’t train models on your private conversations.
• Anonymised aggregate usage helps us improve the coach — opt-out always available.
• One-click delete on all saved data.

Our principles

• Identity-first. Always.
• Honest answers. Specific, grounded, source-linked. No hedge / fluff / apology.
• No ABA, ever. Not in the training material. Not in the model’s instructions. Not in our recommendations.
• No “consult a professional” deflection. When professional input is needed, the coach says so specifically — what kind of professional, what to ask, why. Not as a get-out-of-question-free card.
• Grounded in lived experience. The training material is written by ND adults, not borrowed from clinical journals.
• Privacy-first. Your questions are yours.

A few things people ask