Asclevor
The medical data layer.
The missing infrastructure for healthcare AI. Structured, cross-linked, versioned medical knowledge — delivered through one clean API. Your data stays GDPR-compliant, your queries stay fast, your product ships on time.
Building a clinical product? Talk to our team →
One API. Every condition.
Asclevor delivers 700+ conditions and 300+ symptoms as a clean REST API — differentials, ICD-10 lookups, severity scores, and cross-linked relationships included. Free tier from day one. SDKs for TypeScript and Python. Docs that don't read like compliance paperwork.
Platform
Infrastructure for medical intelligence.
Asclevor builds the tools and infrastructure for searching, understanding, and working with clinical knowledge — semantic retrieval, a cross-linked medical ontology, and programmatic access through one clean API.
Solutions
Built for the people advancing medicine.
One knowledge layer, shaped to the people who use it.
Technology
The technology behind the intelligence.
Semantic retrieval over published clinical cases, a cross-linked medical ontology, embedding-based ranking, and the data infrastructure to run it at scale — designed as one platform.
- Query
A clinical question, in natural language.
- Embeddings
Queries and cases embedded with bge-small-en-v1.5.
- Semantic retrieval
Similarity search across the indexed case base.
- Ranked results
Structured JSON with similarity scores and source PMIDs.
Research
Research at the frontier of medical intelligence.
Our research focuses on the foundations of medical intelligence — how clinical knowledge is represented, retrieved, and evaluated, and how to measure honestly what these systems can, and cannot, do.
Retrieval
Measuring relevance in clinical case search
What does it mean for a retrieved case to be clinically similar? The criteria results are ranked against — and how we measure them.
Embeddings
Embedding published cases at scale
More than 167,000 patient cases embedded with bge-small-en-v1.5 — throughput, index growth, and what we look for in a model.
Evaluation
Honest evaluation for medical AI
Retrieval systems fail quietly. Limitations documented alongside measured strengths, with work in progress labelled.
The dataset
Building what comes next.
Our mission is to make medical knowledge as accessible to developers as modern infrastructure made computing — open at the core, one problem solved exceptionally well at a time.