NLP Building Blocks

NLP microservices for powerful and loosely-connected NLP systems and pipelines.

Our applications are building blocks for larger NLP systems and services. They provide functions integral to text processing such as text tokenization, language detection, document classification, and entity extraction. Each application is stateless, scalable, and features a REST API for easy integration into NLP pipelines and with other applications.

NLP Flow allows for creating NLP pipelines in Apache NiFi.

Use docker-compose to create the NLP building blocks.
NLP Building Blocks SDKs

Idyl E3 Entity Extraction Engine

Idyl E3 Entity Extraction Engine is capable of extracting named entities from natural language text. Idyl E3 supports user-created entity models, has a REST API, and can be used in an air-gapped network.

Idyl E3 Quick Start |  Source Code

Prose Sentence Detection Engine

Prose Sentence Extraction Engine breaks input text into its individual sentences based on rules or trained models. Many NLP applications operate on individual sentences and Prose provides an efficient means of extracting sentences from input text.

Prose Quick Start  |  Source Code

Renku Language Detection Engine

Renku Language Detection Engine detects the language of input natural language text. Capable of identifying over 100 languages, Renku is designed to be lightweight and efficient. Use Renku in your NLP pipelines to route text based on the text’s language.

Renku Quick Start |  |  Source Code

Sonnet Tokenization Engine

Sonnet Tokenization Engine breaks input text into its individual tokens based on rules or trained models. Many NLP applications require tokenized text as input and Sonnet provides an easy to use means to tokenizing text.

Sonnet Quick Start |  |  Source Code