Designed for Decision Making

Mature decision support model incorporating collective and individual opinions, coupled with machine learning to give domain overview over human activity systems.

Diverse Corpora

We collect and combine diverse large-scale corpora in the with 100s of millions documents, both public and private, covering world news, technology, medical and financial domains.


We connect, cluster, train and extract in more than 15 latinate and non-latinate languages including Chinese (Simplified & Traditional), Russian, Korean, Arabic, Spanish, English, etc.

Realtime Machine + Human Learning (not AI)

Our ML algorithms learn from domain experts in real time, providing instantaneous feedback for users. In this tight loop humans and algorithms learn from each other.

HTML5+ Visualization

We embrace the modern web browser as a powerful application delivery platform, even on mobile. We use cutting edge browser capabilities to offer high-resolution visualizations of very large datasets.

NLP & Entity Extraction

We process natural languages at scale, extracting numerous features, including high-level entities. These entities are disambiguated using trusted ontologies.

Fast Access

Our enriched corpuses are indexed and stored on low latency SSDs in bare-metal servers. Billions of items are available instantaneously, via search, apps, or bulk ML processing

Robust Unsupervised Learning

Our clustering technologies are not domain specific and operate at scale. This gives robust overviews of any data set.

Usable Apps

Our applications are powerful, yet accessible to non-technical users. This makes it trivial for experts in various domains to provide training data for our machine learning platform and for experts to gain insight into how the platform shapes their insight.