In contrast to the neat rows of tabular data that drive the modern world, there is an ocean of content in email, chat logs, webpages and PDFs that requires transformation into structured data to be actionable.
Our expertise in natural language processing (NLP) and document analysis can help you to expand your data collection and analysis into all aspects of the digital landscape.
We have created technology that parses PDFs and webpages into structured documents, extracting paragraphs, images and tabular data.
We have created an efficient and scalable text processing pipeline that breaks text into annotated parse trees with likely and inferred semantic relationships.
Our text pipeline identifies named entities and correlates them against a graph based topic/entity model. Relationships between entities are stored in an entity domain graph, with consideration for time based sequencing.
We can analyze text for sentiment and subjectivity and identify intent from questions.
Our expertise with machine learning can create custom predictive models for your domain. Such as using recurrent neural networks to predict outcomes from a sequence of time based text records.