At the core of our solution are three fundamental principles:


To be efficiently leveraged by either a human analyst or an AI algorithms, information and knowledge must be broken down to its smallest data point (either a node, a connection, or an attribute of either a node or a connection)

To support this principle we developed a network graph architecture where data and information are stored in either a node or a connection. Such architecture is the most efficient both from an analytical and a development perspective.

In addition to our data architecture, we also apply the deconstructivist principles to our modular system architecture.  The whole system is made up of thousands of small pieces of code (hardcoded or machine learnt) each doing its own process. The advantage of such a modular approach is that we can multiple development tracks working in parallel to bring new functionalities to the analytical interface.


Human/machine meshing

to develop AI intelligence forecasting, breaking down information and knowledge into data points that can be leveraged by AI algorithms is not enough. For these algorithms to have any chance of producing intelligence insights and forecasts they will need a lot of human inputs and feedback.

QIIK is designed to maximize every human input, not only to improve the quality of its data but also for the purpose of training machine learning algorithms.

In addition, the system is designed to be fully transparent and accountable. For every data point stored in QIIK, the system will also record the source(s) the automated algorithms had found the information in.  This way, the human user can quickly assess the accuracy of the information.

Finally, human inputs are essential to building the analytical overlay necessary to process the raw factual data.

To support such complex requirements we plan to integrate blockchain technology to track all changes made to every data point stored in the system.


Epistemologically, ontologically and method agnostic platform

At QI we firmly believe that human affairs are too complex to be reduced to simple models or theories. Being critical thinkers at heart, we don’t believe our system should define our clients epistemological and ontological frameworks or method.

Such an agnostic approach is all the more important in the field of intelligence where competing analysis such as red teaming is at the core of intelligence production.

As a result, we have designed our analytical platform to support any agenda or theoretical approach. By differentiating factual information from the analytical overlay, the system allows different users to define their analytical parameters and analytical grouping on the fly, using QIIK internal logic or user-defined logic.