Fast. Flexible. Intelligent. Scalable. Secure.
ConceptualEyes is the convergence of many technologies
Natural Language Processing
Reports, emails, documents, surveys, academic publications that make up 70% of enterprise content are unstructured text. ConceptualEyes utilizes the best-in-class methods and tools for parsing and ingesting textual content and exploits innovative algorithms to construct vocabulary, organize relationships and model topics to create comprehensive knowledge graphs.
Scalable Graph Analytics
Knowledge graphs are a collection of sentences that contain a subject, a predicate and an object. At the foundation of the ConceptualEyes platform is a scalable graph store that hosts massive-sized knowledge graphs. On top of these knowledge graphs, we have deployed unique and differentiated software that is able to execute ad-hoc queries and perform graph-theoretic operations. Not only is the graph store capable of handling hundreds of terabytes of data, it is also high-performant - returning results that guarantee interactivity.
Deep Learning is used in many parts of the ConceptualEyes platform - particularly for the natural language processing tasks of identifying parts-of-speech, named-entity recognition and entity resolution. While we leverage pre-existing models for these tasks, the ConceptualEyes team has in addition designed and trained a proprietary language model for writing sentences that rank high on semantic, statistical and logical value.
Deep learning works well when there are enough examples to learn from. With unstructured text data across domains such as healthcare or finance, labels are expensive, cumbersome to collect, or not readily available. ConceptualEyes leverages cognitive heuristics such as saliency and attention to self-organize and seed labels for the deep learning problems. These heuristics are critical in navigating knowledge graphs in an intelligent and interpretable manner.
Real-time Machine Learning
Deep learning, cognitive heuristics, natural language processing and graph analytics when assembled into a workflow help us search for the "what-is" (i.e. a data retrieval task). The real-time machine learning component of ConceptualEyes implements the ability to build models based on query-context at the time of search. The output of the machine learning component extrapolates based on observed textual and conceptual patterns to come up with futuristic "what-if" suggestions.
We know that privacy, security and user-level customizability are important requirements to our customers. Today, we deliver ConceptualEyes as a packaged solution that wraps all the aforementioned components. Our solution can be deployed both on cloud-hosted and self-hosted infrastructure with support for Docker containers.
ConceptualEyes is architected for flexibility, security, scalability and performance.
ConceptualEyes is best described as an artificial intelligence system that discovers and ranks meaningful associations by reasoning with unstructured text documents. As a software application, it is a suite of scalable algorithms for semantic, logical and statistical reasoning with Big Data (i.e., data stored in databases as well as unstructured data in documents). It uses a high-performance graph analytics engine that manages thousand-fold larger datasets and operates hundred-fold faster than traditional graph databases.
Over the last few years, this platform has evolved into a futuristic next-generation knowledge-discovery framework that is:
- Knowledge nurturing - evolves seamlessly with newer knowledge and data
- Smart and curious - using natural language processing, intelligent data parsing and harmonization, information-foraging and reasoning algorithms to digest content
- Discovery enabling - interfaces computers and artificial intelligence with what they do best to help subject-matter-experts do their best
Read more about the different technologies underlying the engine from the links provided below.