Cognizer
Redefining the Future of Knowledge Management

At an investor meeting, Jack Porter, the President and CEO of Cognizer, challenged his co-attendees to test the intelligence of Cognizer’s solution, “Corporate Brain,” by asking questions to it. “What is the meaning of life?” asked one among the attendees. The Cognizer Corporate Brain instantly replied, “42,” a popular answer inspired from Douglas Adams’ science-fiction novel, The Hitchhiker’s Guide to the Galaxy. Dubbed as the world’s first AI-powered Corporate Brain, Cognizer hardly took a second to answer many such questions that came its way. “The crowd was astonished to see the potential of Cognizer. The response from them was overwhelming,” Porter recalls.

“What spurred the inception of Cognizer’s Corporate Brain is the need for enterprises to have easy and quick access to the right business information that will help them stay ahead of the pack,” mentions Porter. As today’s business landscape is witnessing an “information overload”—which will spiral into the 2020s—many organizations and leaders are already burdened with humongous amounts of data. This data is often scattered across employees’ systems, storage systems, and more. Searching for a file or specific information in such data silos often impedes enterprises’ effort to access and leverage knowledge. Having worked as the CEO of multiple companies, Porter has firsthand experience of these challenges.

“Consider retaining relevant data from around 500 emails daily and recalling them to fulfill business commitments,” he says. The massive influx of information from various sources, including meetings and conversations, overwhelmingly surpasses the brain’s capability to store and process it. In such a scenario, companies need powerful solutions to retain and disseminate business knowledge from multiple sources and leverage them as a core part of their strategy and corporate DNA. Cognizer is built to perform this through an innovative approach that will forever change how companies manage knowledge within their organization.

“Cognizer is not designed to be smarter than you, but to make you much smarter,” says Porter. It is a new form of AI that promises a “superpower” for enterprises to know more, react faster, be more agile, and outsmart the competition. “Corporate Brain will help businesses understand the past, respond to the present, and predict the future,” Porter adds. Unlike any other virtual assistant tools in the market, Cognizer augments the user’s ability to “know” rather than merely spend time in “searching.” The solution makes people smarter by continually providing them with comprehensive, personalized, and real-time information.

Powering the “Superpower”

Cognizer’s Brain can read corporate contents from email, file systems, Box.com, intranets, websites, and even enterprise systems like CRM, HRMS, and ERP. As the solution reads all content, it finds all named entities (people, places, locations, technologies, etc.) and the facts about those entities. It puts these into a massive fact store that is based on a high-speed graph database. The solution also ingests metadata like tags, attributes, and properties. This information is indexed using an advanced temporal-based graph system that organizes data into a comprehensive ontology and produces a sophisticated, fast, accurate, and powerful knowledge base for instant retrieval in the form of “Fact-Tiles.”

Users can ask The Brain any question that is anywhere in the fact store, and it will provide the answers with sub-second response. It can also listen to the user and provide information about the conversation in real-time as the words are spoken.

Corporate Brain will help businesses understand the past, respond to the present, and predict the future

What propels the Cognizer Corporate Brain is its Natural Language Intelligence (NLI), also called “The Cognizer Engine” that turns siloed data into information and then into valuable insights. The engine scrolls through vast amounts of corporate content—be it structured, unstructured, or semi-structured—and curates the data to build semantic graphs for outputting instant insights. Cognizer’s NLI is powered by deep-learning meta-models and advanced linguistic machines to process structured data, sophisticated neural networks for unstructured data, as well as deep sentiment analytics engines. It comprehensively preprocesses source content to enhance the data readability and mitigates chronic NLP challenges pertaining to spelling, slang, bad grammar, icons, as well as contractions.

With a continuously adaptive approach to enhance their solution, Cognizer incorporates HyperFraming into their system; they extend FrameNet, a powerful computational linguistic database, to their natural language processing (NLP). This database, coupled with deep learning enables the firm to increase the number of semantic frames that FrameNet provides. Cognizer’s HyperFrames are capable of understanding the subtleties of language, such as sarcasm, irony, and slang. The company also integrates its deep learning stack with XLNet, a generalized autoregressive pre-training method for NLP, to improve accuracy.

"Cognizer is not designed to be smarter than you, but to make you much smarter"

On top of that, Cognizer’s Corporate Brain has a direct connection to Wolfram Alpha’s computation engine and can do Ph.D. level math instantly. Its robust preconfigured math engine solves arithmetic, algebra, calculus, and differential equations by using any fact it knows. For a better understanding of Cognizer’s mathematical capabilities, consider how a shipping company used the solution to optimize their capacity. The Brain was able to calculate the number of items that can be loaded into a container after capturing the size of the product and the container. In another instance, the platform instantaneously performed the math and answered the question, “How many quarters does the moon take to cover the surface area of earth?”

The Technology Neurons Firing the Corporate Brain

Cognizer has a proprietary multi-tier memory system, which Porter describes as “the brain of the Corporate Brain.” It stores all types of information in the long-term memory, which is based on a triple fact store and sophisticated graph database. The Brain also comes pre-learned with informational corpora, including Wikipedia, Wikibase, DBpedia, Wikidata, the CIA World Factbook, Freebase, and WordNet. The memory includes 4.58 million facts about 1,445,000 people, 241,000 organizations, 735,000 places, 411,000 creative works, 251,000 species, and 6,000 diseases, generated by reading over 200 million pages of material.

Additionally, by leveraging “short-term memory,” which uses multiple sliding windows-like approaches to carry out the analysis, The Brain can understand the context of data-be it in a sentence, the paragraph, the topic of discussion, or the content as a whole.

For instance, consider the platform reading a bunch of statements: “Jack Porter is the CEO of Cognizer. His wife’s name is Susie. Their kids’ names are Becky and Katie. Becky is 27, and Katie is 24.” While generic NLP machines fail even to understand what Jack and Porter are, Cognizer can answer questions like “What is the name of Susie’s husband?” which is nowhere present in the statement.

Modes to tap-in to the Corporate Brain

Businesses can run the Cognizer Corporate Brain on the web, mobile devices, and soon via powerful augmented reality interfaces. “We believe that Cognizer will be the “Killer App” for augmented reality,” says Porter.

Such a level of integration allows organizations to gain accurate, real-time knowledge without crossing security and privacy boundaries. The platform will never return data to an unauthorized user. On the contrary, when a user leaves the organization, The Brain can retain the knowledge he gained and place it at the disposal of the management or the newly hired replacement.

In time, as insights built-up, Cognizer combines all this information with its fact directory in the knowledge base to assist businesses in their operations by providing answers and computing forecasts. Besides this, the firm delivers a knowledge dashboard, which displays metrics about the accumulated knowledge, including how the information is growing over time, how value is being added, and who the top contributors are. By analyzing these contents, it can address all action requests, such as tasks assignments, new contracts, and follow-up meetings, and process these through the correct workflow.

Multiple Models to Enhance the Delivery

Cognizer packs its all-in-one knowledge management solution in three modules that can be readily leveraged by enterprises— BrainAssistant, BrainWaves, and MyBrain. BrainAssistant is a Q&A system, which can be used as a mobile application or with collaboration tools such as Slack, Skype, and Teams. With the help of its unique real-time research assistant BrainWave, The Corporate Brain listens to the conversation and provides instant knowledge on the topics being discussed. The MyBrain application, a powerful time management assistant, conducts analytics on personal calendars and helps employees optimize the use of time and even prepare for a meeting.

Cognizer promises to revolutionize how enterprises retain and disseminate business knowledge throughout their ecosystem. For example, consider the pharmaceutical sector. Companies in this sector often fail to communicate outcomes of clinical trials with their branches, which can lead to serious issues. Cognizer can wrap the whole enterprise (across all branches or units) with a readily accessible, virtual layer of knowledge that can distill crucial information proactively across the teams and departments. Along similar lines, entities in the IT and Legal space can also benefit massively from the solution.

To strengthen Cognizer’s edge in the knowledge management space, the company’s innovation lab is developing three applications to be released next year. One among them is Relationship Intelligence that will give insights across users’ relationship touchpoints, including calls, documents, meetings with people, and events. Other upcoming applications include Project Intelligence and Location Intelligence.

Going forward, Cognizer is planning to integrate the solution into other document repositories such as Google Drive and OneDrive to expand its customer base. The company also has plans to focus more on the prediction model by leveraging the newer and powerful Hierarchical Temporal Memory. “Through this, we are planning to disrupt the legacy models and go beyond deep learning, which is the fuel of our solutions now,” concludes Porter.