Savana is your
AI Scientific partner
for RWE generation.

Through our AI-powered NLP technology, we have already structured more than 5 billion clinical documents in 6 languages and in more than 10 different countries.

Savana is your
AI Scientific partner
for RWE generation.

80% of all clinical data are found within the free text(1).

(1) JAMA. 2013;309(13):1351-1352. doi:10.1001/jama.2013.393

But often remains inaccessible for clinical and research purposes.

Trained with more than
5 Billion medical documents.

Input Data Types

We can process .PDF, .word and .txt and all the free text alocated in Electronic Medical Records.


Vast experience in interconnecting with multiple EHR, RIS, LIS and other information systems.


Unique enriched clinical terminologies developed by Savana during more than 10 years.

Output data

There are several formats to send back your structured data. We can work with standards as OMOP, FHIR and openEHR.

Why is NLP so important?

This example is an internal study performed by Savana with real Hospitals data.

Clinical NLP Models as a service.

Apply a set of general models to detect clinical variables and their attributes.

01. Clean free text from unwanted artifacts (e.g. copy-paste).

02. Associate identifiers from a standard terminology to clinical entities.

03. Train specific models to increase non-standard entity detection.

NLP Evaluation

methodology published.

This is the first step for AI + RWE:

A KAM will contact you to schedule a NLP as a service demo:

Complete the info, and a KAM will contact you ASAP:

Want to use it?:

Start with your proposed AI + RWE use case: