Multilingual & Scalable Data Extraction

Next Generation Natural Language Processing Software


An enterprise scalable solution to extraction

Identify over 3 dozen entity types
Identify 500+ relationship types
Multilingual platform in 200+ languages
Sentiment metrics and salience scores provide insight on the emotion and importance of the text


Create new entity types, relationships and customer-specific lexicons to fine-tune results

Provided utilities: document retrieval, licensed-based routing, tokenizing, rule checking, output formatting

This ultimate extraction solution is completely turnkey, allowing administrators to be up and running in as little as 10 minutes. Rosoka's API-driven engine allow users access through a Java API or REST web service and outputs results in XML, JSON or POJO.

The Rosoka Series 7 Server package delivers a Manager & Worker based solution that supports custom processing workflows and load-balancing for true scalability. The Rosoka Server offering includes all of the Rosoka multilingual entity, relationship, and geospatial extraction, and text analytics features. Like all Rosoka solutions, the extraction rules can be modified to support industry-specific needs.
The Rosoka SDK is more than just the Rosoka Extraction Engine. The SDK includes all the tools necessary to integrate the Rosoka Extraction Engine into any third-party solution. The Rosoka SDK for Developers includes everything needed to develop an integrated solution that offers Rooska's multilingual unstructured analytics capabilities.
The SDK package includes:
- 1 Rosoka Extraction Engine (non-production, development license)
- Java API libraries
- REST API libraries
- Python API libraries
- Javascript GUI libraries
- REST Microservice
- Documentation
- Code samples
- Utilities (i.e. IID creator/decoder)

Rosoka Studio provides the data scientist with an easy-to-use User Interface to modify and create: 

- Entity types
- Relationship definitions
- Customized lexicons
- Extraction rules
- Quality control with built-in regression testing
- LxBases for simple import and export to obtain customized extraction results.


Rosoka Text Analytics provides IBM i2 users with advanced text analytics features that turbocharge the speed of analysis and drives better decisions. From the moment the analyst imports documents, Rosoka Text Analytics is ready to process the text from virtually any source and language. The analyst is presented with an intuitive interface that presents all the important entities and links, categorized by type, to make focused analysis a breeze. The application was designed to allow the analyst to remain in control and apply their expert knowledge to build a finished work product by giving them the ability to hand curate extraction results in the integrated document viewer. Analysts can send entities of interest to the chart and with a single click expose the subtle but important relationships across a large collection of documents. Analysts can also instantly review the source document(s) that contain the entities or links of interest to audit the results. The final work product is always controlled by the analyst.


Rosoka GeoGravy uses linguistic context to provide accurate geocoordinate tagging for Places and Facilities across various formats.


Portable. Scalable. Multilingual.

Deploy across any language, platform, application, device, or cloud.

Up and running in as little as 10 minutes.

Schedule a Demo