TL;DR
AI is not replacing your therapists. It can’t deliver a session, build rapport with a kid, or make the in-the-room judgment calls a trained BCBA makes a hundred times an hour. What it can do is give a trainee a place to rehearse those judgment calls before they’re on the line. That’s the role Kipr plays — a practice partner, not a practitioner.
Let’s draw a clean line
We’ll save you the marketing dance. Here’s what AI in this field can and cannot do, in 2026, the way we see it.
What AI can’t do:
- Deliver an ABA session
- Build a real therapeutic relationship
- Replace a BCBA’s clinical judgment
- Make ethical decisions about a treatment plan
- Read a real child the way a trained clinician reads a real child
What AI can do, well:
- Hold a consistent persona across many practice reps
- Respond to a trainee’s antecedent setup, prompts, and pacing in plausible ways
- Generate the same scenario at different difficulties on demand
- Give a trainee a dignified place to fail and try again
- Free a supervisor’s time from rehearsal so it can be spent on judgment
Kipr lives entirely in the second list.
Why the “replacement” frame is the wrong one
When people hear AI for therapy, the first instinct is often to imagine a chatbot delivering treatment to a child. That’s a real product category in adjacent fields (mental health support, coaching), and it raises serious questions there. It is not what we are doing.
ABA is delivered by people. The BACB credential exists for a reason. The therapeutic alliance between a BCBA, an RBT, a learner, and a caregiver is irreducibly human. We are not interested in replacing it, and frankly, we don’t think the field should tolerate anyone who is.
What we are interested in is the bottleneck no one in the field has solved: how do you give a new behavior analyst the reps they need to be ready for that human-to-human work?
How Kipr works in practice
Kipr is a practice platform for behavior analysts and behavior technicians. The shape of it:
1. Practitioner-designed personas. Every persona — a learner, a caregiver, a paired-stim setup — is built with input from BCBAs and clinicians who run real sessions. Behaviors, antecedents, reinforcer profiles are grounded in clinical realism, not Hollywood drama.
2. Scenario library by skill. Trainees don’t pick random scenarios. They pick the skill they need to work on — paired-stim assessment, escalation recovery, NET pivot, caregiver redirection — and run reps inside scenarios designed to surface that skill.
3. Feedback that points at the moment. After each rep, the trainee gets a transcript or recording with marked moments — prompt missed at 1:42, recovery delayed by 6 seconds, reinforcement timing improved on second trial. The feedback is concrete, not motivational.
4. Supervisor view. Supervisors see what their trainees have practiced, where they’ve plateaued, and which scenarios to prioritize in the next live supervision meeting. The platform makes the supervisor’s hour more diagnostic.
5. Privacy and ethics by design. No real client data is uploaded into the simulator. Personas are synthetic. Trainee data stays in the trainee’s account and the supervisor’s view — nothing is repurposed.
What it’s not
- It’s not a CEU mill. (We are pursuing BACB Type 2 CEU approval for specific modules, and we’ll be transparent about which content carries CE credit when that lands.)
- It’s not a replacement for fieldwork hours.
- It’s not a chatbot wrapper. The persona behavior, scenario design, and feedback layer are the product.
Who this is for
We built Kipr with three groups in mind:
- University ABA programs that want their students to graduate with rehearsed reps, not just coursework.
- ABA clinics that are onboarding RBTs and BCaBAs and want a faster, safer path from 40-hour course to clinic floor.
- BCBA candidates and supervisors who want to make the most of the supervision hours they already pay for.
If one of those is you and the line we’re drawing — practice partner, not practitioner — sounds like the line you’d draw too, we’d love to talk.
Join the waitlist for early access →
New here? Start with the case for simulation in BCBA supervision, or the five skill gaps new RBTs face on day one.