NeuroTalk is entering the U.S. market. We're recruiting ASHA-certified SLPs to pilot our AI-powered articulation and cognitive rehab platform β and tell us what works in real American clinics, schools, and SNFs.
These aren't abstract problems. They're the daily reality in 13,000+ special education schools, 15,600+ nursing homes, and countless outpatient clinics across the country.
More than 1 million Americans are currently on speech therapy waitlists. In SNFs and school districts, SLP caseloads are stretched thin β visits often drop to weekly or less.
Between sessions, patients have no structured, AI-guided practice. Progress stalls without daily reinforcement β and you're left repeating the same goals month after month.
Even with insurance, out-of-pocket costs average $150β$350 per session. For stroke survivors, Parkinson's patients, and children with developmental delays, long-term therapy becomes financially unsustainable.
Sources: ASHA, WHO, Research and Markets (2024β2025); CDC & NIH 2022 reports. Dementia patients in the U.S. rose from 323,271 (2020) to 418,167 (2023); stroke patients from 95,354 to 103,866 in the same period.
NeuroTalk augments your clinical workflow β automating drill practice, tracking phoneme-level progress, and keeping patients engaged between visits. It aligns with IEP goals, supports LSVT protocols, and gives you data for progress reports.
Standard speech recognition (Google, Siri) drops below 45% on dysarthric speech. Our LLM-tuned model β trained on 140M+ disordered-speech data points β achieves 89.58% accuracy on impaired articulation, nearly double the industry standard.
The system auto-generates a tailored therapy curriculum for every session based on real-time error patterns. No manual configuration required β but fully aligned with your IEP or Plan of Care goals.
Systematically reduces support from full model β visual cue β syllable prompt β spontaneous speech. Prevents tablet dependency and promotes generalization β backed by the behavioral principles you already use.
Track training frequency, accuracy trends, error phoneme clusters, and home compliance in a web-based dashboard. Use the data to inform your next session β or your next Medicaid / insurance progress report.
Every module is grounded in SLP methodology β just deployed at the speed and consistency that only AI can deliver.
LSVT-inspired vocal intensity training: sustained vowels β words β sentences. Designed for Parkinson's dysarthria, vocal hypofunction, and any patient needing increased vocal loudness.
Paired-stimulus articulation therapy targeting high-frequency English phonemes like /r/-/w/, /s/-/ΞΈ/, and /Κ/-/tΚ/. Discrimination + production drills in one flow.
Type any target word or sentence directly into the app. The AI analyzes phoneme-by-phoneme accuracy in real time. Perfect for aligning with IEP objectives or individualized Plan of Care goals.
Place-of-articulation Γ frequency auditory discrimination using non-words. Specialized for cochlear implant recipients, hearing aid users, and aural rehabilitation cases.
AI auto-generates a new drill set each day based on cumulative error data. Your patients get fresh, targeted practice without you writing a new worksheet every night.
Save high-priority targets for rapid session setup and maintenance / generalization probes. One tap, and the patient is practicing their exact IEP or therapy goals.
Pilot data from Korea's national certification lab (KCL) β an independent, ISO-accredited body. U.S. FDA-tracked trials are on the roadmap for 2029β2030.
20 subjects, 4-week training (1 hr/week). Percentage of Correct Consonants improved by +6.9 percentage points. Wilcoxon signed-rank test: p<0.001, Cohen's d = 0.52. Targets exceeded by 130%.
KCL Independent Certification β Korea's National Lab14 seniors (ages 60β80+) trained 30+ min/week for 4 weeks. Mean K-MoCA improvement +1.71 points (p=0.028). 71.4% showed cognitive gains; 21.4% converted from MCI to normal range.
KCL-Supervised Clinical TrialN=35 adults ages 60β80+ scored 90+ across all 5 usability dimensions. Even digital-naΓ―ve elderly users found NeuroTalk intuitive β suggesting strong accessibility for your geriatric caseload.
KCL + In-House Validation (N=38)FDA 510(k) or De Novo submission targeted for 2029, with SaMD Class II clearance expected by 2030. MFDS (Korea FDA) Class 2 DTx approval for post-stroke dysarthria planned for 2027.
Company IR & Regulatory Deck
We're building this tool with SLPs β not for them.
Your frontline perspective on what works (and what doesn't) in American schools, SNFs, and clinics is the most valuable input we can get.
Full NeuroTalk Solo access for the interview period β no hardware, no credit card
Compensation for your time and clinical expertise
Summary of anonymized findings + first access to U.S. product updates
We'll reach out within 2 business days. Interviews available in-person (Seattle / San Francisco) or remote via Zoom. All data handled confidentially.