therapy4speech.aiTherapy is once a week. Recovery needs every day.
Speech and language disorders are common, treatable – and chronically under-served. The gap is not the therapist's skill; it is the days between sessions.
Recovery is a dose problem
What the evidence says works – and what patients actually receive.
- Trials with about 8.8 hours of therapy a week showed significant effects; trials with about 2 hours a week did not (Bhogal et al. 2003)
- Best outcomes at 20–50 total hours, 3–5+ days a week, with prescribed home practice – more intensity than usual clinical services provide (RELEASE Collaborators 2022, 959 patients, 25 trials)
- For children with speech sound disorders, three sessions a week beat one with the same total number of sessions (Allen 2013)
- 69 % of therapists can only deliver therapy once a week (Hegarty 2018)
How it works
Prescribed by your therapist · coached by your camera · practised at home. Two AI layers, one clinician in charge.
Therapist prescribes
Sets the goals. Generative AI drafts a personalised exercise plan; the therapist edits and approves it in one click.
Patient plays
Front camera on. An on-screen guide shows the movement or the word; the patient follows it – like a dance game – every day, for as long as the therapist prescribes.
The phone coaches & scores
Speech recognition and facial-landmark analysis score every attempt: points, streaks, levels. Designed for on-device processing and data minimisation.
AI reports back – the therapist decides
Plain-language progress for the family, a clinical summary for the therapist, who adjusts the plan. Low confidence or regression always escalates to a human.
See it in action
Three screens from the prototype – the plan from your therapist, live gamified feedback, and the report that goes back to the practice.
Home
Today's plan from your therapist – a few exercises, a streak, a start button.
Exercises
Front camera and microphone on: the guide shows the movement, you copy it, every attempt is scored live.
Report
Your session summary – shared with your therapist, who adjusts the next plan.
Prototype screens, September 2026. Video and audio are processed for scoring; raw video is not stored by default.
Amplify therapists – don't replace them
Machine learning evaluates each patient's trajectory – improving, plateauing, regressing – and proposes the next step under explicit clinical rules. Generative AI creates the exercise for that need, grounded in a curated clinical knowledge base. The therapist reviews, adapts and keeps clinical authority.
Target: therapist review time per evaluation from 25 to 10 minutes – so one therapist can follow a full caseload, every day.
Built by people who live the problem
Five countries, one mission. Our CPO has stuttered all his life; our CTO is married to a speech therapist; a third team member has close family experience with stutter.
If someone you love needed speech therapy – which therapist would you choose?
Pilot programme 2027 – we are looking for partner clinics: two pilot sites, a clinical advisory board and therapists who want to shape the product. Pricing on request.