Clinician-validated, HIPAA-safe, zero PHI. 25 document types across four disciplines — ABA, psychotherapy, speech-language pathology, and occupational therapy. Schema-validated, audit-trailed, ready to train.
25
Document Types
163+
PHI Detection Rules
10-Gate
Quality Pipeline
SHA-256
Provenance Hash
BCBA-Designed
Pipeline architected by BCBA-candidate clinician with master's in behavior analysis. Dual-critic AI validated against BCBA-D clinical standards.
Zero PHI
163+ rule VLayer PHI scanner — no real patient data, ever
Full Audit Trail
VLayer compliance passport with SHA-256 provenance, quality gates, and chain-of-custody for every record
VLayer Scanned
Every record passes through automated PHI detection and synthetic validation via VLayer Learn more →
Every record passes through four verification stages before release.
Raw document enters the generation pipeline
163+ rules verify zero real patient data
Clinical plausibility and schema conformance
SHA-256 hash sealed in compliance passport
Four clinical disciplines. Every document type used in real practice — schema-validated with Pydantic and clinically plausible.
Session documentation — subjective, objective, assessment, plan
Antecedent-Behavior-Consequence recording sheets
SMART goals with baselines, targets, and measurement criteria
Authorization requests with clinical rationale per insurer
Behavior intervention plans for emergency situations
BCBA supervision session documentation
Treatment completion reports with outcome data
Periodic progress updates with quantitative metrics
Raw trial-by-trial and interval recording data
VB-MAPP, ABLLS-R, and standardized assessments
Individual therapy sessions — CBT, DBT, EMDR modalities
Initial psychiatric evaluation with HPI and MSE
Full 12-domain mental status examination
Suicide prevention — Stanley-Brown safety planning
Group sessions — process, psychoeducational, skills training
Family therapy with dynamics and interventions
Cognitive, emotional, and personality assessment reports
SUD evaluation with DSM-5 criteria and staging
Initial speech-language evaluation with standardized tests
Treatment session with cueing hierarchy and accuracy data
Periodic progress with goal tracking and justification
Goals, objectives, frequency, and approaches
Initial OT eval — motor, sensory, self-care, VMI
Treatment session with performance data and assistance level
Functional goals, modalities, and home program
Our generation pipeline implements the four principles of Google Research's Simula framework, specialized for behavioral health. The result: coverage of the long tail, not just the easy cases.
Every dataset samples from a 113-node clinical taxonomy spanning behaviors, interventions, assessments, comorbidities, settings, phases, and barriers. Coverage ratio reported per dataset.
Every document passes independent review by two AI clinicians — one checking accuracy, one actively looking for errors. A third reviewer reconciles. Breaks the sycophancy bias that plagues single-LLM validation.
Each document is scored 1-10 on clinical complexity and calibrated with Elo ratings through pairwise comparison. Your team filters for edge cases the simple way: complexity_elo > 1700
Every document ships with a complete audit trail: taxonomy path sampled, complexity scores, both critic verdicts, HIPAA scan results, model versions. Auditable by your legal and ML teams.
One-time purchase. You own the data. Includes full quality reports, documentation, and schema exports.
Standard
$25,000
1,000 documents
Complexity 1-6. Coverage across common clinical scenarios.
Premium Edge Case
$85,000
500 documents
3.4× per-document rate — justified by the edge cases your model fails on
Complexity 7-10. Rare clinical combinations that break models in production.
Enterprise
Custom
Custom volume
Multi-discipline datasets, custom document types, SLA, dedicated support.
Everything your technical, legal, and compliance teams need to evaluate and approve SynthABA.
Executive overview — what SynthABA is, 25 document types, pricing, and delivery format.
Read →Complete field-level reference for all 25 document types, enums, CPT codes, and ICD-10 codes.
Read →API reference, code examples, batch structure, schema validation, and delivery formats.
Read →10-gate pipeline results, PHI detection, demographic distributions, and inter-rater reliability.
Read →Demographic distributions vs. CDC prevalence, known limitations, and mitigation strategies.
Read →Legal analysis: why SynthABA data is not PHI under HIPAA Safe Harbor (45 CFR 164.514).
Read →Template DUA covering permitted uses, restrictions, security requirements, and termination.
Read →Formal certification that the dataset contains zero real patient data. Signable.
Read →Enter your work email and we'll send you a free sample dataset — 5 records across document types so you can evaluate schema quality, clinical plausibility, and format compatibility.