Papers indexed
10,463+
Seven research sources, vectorized and curated for meaning-level retrieval.
Clinical trial discovery
Semantic expansion for clinical trial discovery.
Complete the trial landscape before you model it.
Exact-match search is a structural limitation of trial registration. RevasserKernel expands the condition first, then queries ClinicalTrials.gov across the full semantic set — seven tiers of curated terminology built from real literature.
Conditions curated
65
Hand-curated condition graphs spanning rare disease, neuropsychiatry, oncology, and adjacent clinical territory.
Expansion tiers
7
Exact → parent → region → system → related → comorbid → therapeutic.
Infrastructure cost
$0/mo
Cloudflare edge architecture. No monthly hosting bill for deployed portals.
Semantic expansion
One query. Seven semantic expansion tiers.
The engine does not guess synonyms. It moves across a curated graph built from real literature, real condition relationships, and explicit tier logic.
Surface trials keyword search leaves behind by widening from the registered term to the surrounding clinical landscape: broader categories, anatomical variants, organ-system involvement, related conditions, common comorbidities, and therapeutic pathways.
T1
Exact match
Primary condition name as registered in the trial record.
T2
Parent condition
Broader diagnostic category and adjacent coding parents.
T3
Anatomical region
Regional and structural variants used in registration language.
T4
System involvement
Physiological systems implicated by the condition.
T5
Related conditions
Clinically adjacent conditions with shared mechanisms or presentations.
T6
Common comorbidities
Frequent co-occurring conditions visible in the literature.
T7
Therapeutic targets
Mechanism-linked interventions that cross condition boundaries.
Exact-match results
408
What a narrow PTSD query yields on its own.
Expanded results
3,904
What the seven-tier semantic graph surfaces.
Lift
+857%
Measured proof that the hidden landscape is large enough to matter.
Products
Condition-specific trial intelligence for the organizations building access.
The same semantic core ships two ways: as an API for teams building on top of it, and as a full research portal for communities that need a real home for the work.
Developers · Funds · CROs
TrialFinder API
Semantic expansion for clinical trial discovery. Submit a condition and retrieve the full expanded trial landscape — deduplicated, ranked, and exportable.
- Seven-tier semantic expansion graph built from curated condition mappings.
- Deduplicated NCT IDs with ranking by proximity, phase, and enrollment context.
- JSON and CSV export, batch workflows, and a live public demo.
- Cloudflare edge delivery with sub-second response characteristics.
PTSD proof case: 408 exact-match trials became 3,904 after semantic expansion.
Free · $99/mo Pro · $499/mo Enterprise
Foundations · Patient advocacy
Research portal
A branded research portal for rare-disease communities. Curated papers, semantic search, collections, exports, and deployment in 48 hours.
- Condition-specific paper index seeded from seven research sources.
- Search by meaning, not just keywords, across the full condition corpus.
- Domain classification, collection curation, exports, and your own branding.
- $0/month infrastructure once deployed — no surprise hosting bill later.
The portal runs on the same retrieval layer already powering the live RevasserKernel corpus and TrialFinder.
$997 DIY · $7,500/yr Managed · Custom
Why this exists
The core problem is terminology drift, not missing data.
Patients, researchers, and operators are all penalized by the same structural failure: the same condition gets registered under different names, so relevant work disappears from exact-match search.
Patients searching for clinical trials face a language problem. A trial for post-traumatic stress may never appear in a search for PTSD unless someone mapped that relationship first. The same failure repeats across rare disease, neuropsychiatry, and every edge case where registration language drifts from community language.
RevasserKernel maps those relationships directly. One query becomes a structured seven-tier search anchored in literature, condition graphs, and operator reality. The graph is the product; the interface is where the proof becomes legible.
Institutional close
Start with one condition. Build the proof from there.
Run the demo. Read the prospectus. If your organization needs a portal, a report surface, or direct API access, move straight to contact.