Why SuperTruth
Data quality tells you what’s wrong. Data trust tells you whether to act on it.
Every health AI team already has data quality tools. None of them answer the question that matters before a model trains or a clinical decision is made: can I trust this record?
The difference
Quality checks the format. Trust checks the foundation.
Data quality
Is the record complete and correctly formatted? Does it pass validation rules?
Answers: What does this record contain?
Data integrity
Has this record been tampered with? Is the chain of custody intact?
Answers: Has this record been corrupted?
Data trust: SuperTruth
Where did this come from? Was consent obtained? Is it current? Does it agree with other sources? Has it been validated against outcomes?
Answers: Should I act on this record?
Feature comparison
What SuperTruth does that data quality tools don’t.
| Capability | Traditional data quality | SuperTruth DTI™ |
|---|---|---|
| Field completeness check | ||
| Format and schema validation | ||
| Duplicate detection | ||
| Source provenance tracking | ||
| Consent verification | ||
| Recency scoring by modality | ||
| Cross-source concordance | ||
| Outcome validation | ||
| Single composite trust score (0 to 100) | ||
| Score travels with the record permanently | ||
| Configurable weights by audience | ||
| Regulatory-grade (FDA, HIPAA, GDPR) | Partial |
Named comparison
How the market compares on what matters.
The question is not whether your AI model can explain its output. The question is whether the data that trained it had any right to be there in the first place.
2019
Year we started building
8
Dimensions scored on every record
2 hours
Review time, down from three weeks (imaware, 105,000 records, published April 2026)
| Company | Data-level trust | Consent | Provenance | Health |
|---|---|---|---|---|
| SuperTruth | ||||
| Seekr | ||||
| Datavant | Partial | Partial | ||
| Scale AI | ||||
| Innovaccer | Partial | |||
| AWS HealthLake | Partial | |||
| Great Expectations / Monte Carlo | ||||
| Komodo Health | Partial |
Based on publicly available product documentation and positioning. SuperTruth does not make claims about competitor products beyond what they publicly describe. The companies named here are not affiliated with SuperTruth and have not endorsed it; their names are trademarks of their owners. The DTI scoring system, the AI Orchestration Engine (AIOE), and the ALDR adaptive learning architecture are patented; DTI™ and Data Trust Index™ are trademarks of SuperTruth Inc.
Head to head
One sentence on each.
SuperTruth vs Seekr
Seekr scores how a model reached its output; it does not score the data the model trained on, and a model that can explain a wrong answer is still wrong.
SuperTruth vs Datavant
Datavant links records across institutions with de-identified tokens; two records can be perfectly linked and completely untrustworthy.
SuperTruth vs Scale AI
Scale AI labels data for training runs; by the time data reaches a labeling pipeline, where it came from and whether it was consented has already been answered, or not.
SuperTruth vs Innovaccer
Innovaccer aggregates records into a health data cloud and uses them as they arrive; a health data cloud is only as trustworthy as the data inside it.
SuperTruth vs AWS HealthLake
AWS HealthLake stores and queries FHIR data; storage is not a trust signal, and no score travels with the record.
SuperTruth vs Great Expectations / Monte Carlo
Great Expectations and Monte Carlo check completeness, format and consistency, three of the eight dimensions at best; schema validation is not consent.
SuperTruth vs Komodo Health
Komodo Health aggregates real-world evidence for analytics; aggregation adds quantity, not a score.
For teams building health AI
Where did this training record actually come from?
You have scraped the internet. You have trained on books, code and scientific literature. The one dataset that matters most for health AI, patient records, lab results, wearable streams and clinical notes, is the one dataset where provenance, consent and recency are never verified before the model sees it.
A clinical note from a CLIA-certified hospital system carries different weight than a self-reported entry from a consumer wellness app. Most health AI training pipelines ingest both without distinction, the model learns equally from both, and the difference shows up in production. DTI™ scores source pedigree and chain of custody for every record before it reaches your training run or your inference pipeline, so your training run knows exactly what it learned from.
The DTI Engine spec →Common questions
Questions we get about the competition.
What is a SuperTruth alternative?
SuperTruth has no direct alternative. It is the only platform that scores health data records across eight trust dimensions before AI models train on them. Data quality tools like Great Expectations check format and completeness. Data networks like Datavant link records. Neither scores for trust.
Is Datavant a data trust platform?
Datavant is a health data tokenization and linking platform. It connects records across institutions using de-identified tokens. It does not score data for trust dimensions like consent, recency, or provenance.
Every competitor starts at the model. SuperTruth starts at the data.
30 minutes. A real sample from your environment. A DTI score you can defend.