SQREEM – The holy Grail of Predictive Analytics
The holy grail of analysis is the ability to distinguish between similarity and causality. Is A caused by B or are A and B caused by C? This has always presented an enormous challenge for conventional statistics. (not due to inadequacy of its practitioners but due to the vastness and complexity of factors surrounding any single point in our universe). A correct answer may simply reside outside a sample of analyzed observations. Or, it may reside in a completely different data structure and domain of measurements.
Specifically, the world of human behavior is simply too complex to be reduced into a collection of correlated data sheets, regardless of how pure, impartial, and proper a statistical analysis may be. At the same time, we have reached the junction where the dynamics of human behavior are as much manifested in the digital domain as they are in the physical. In other words, the data exists, but the challenge resides in conquering its volume, its many sources, and its complexity for any singular process to be applied to a multitude of outcomes.
SQREEM is at the forefront of complex system analysis in a number of specific areas which, when combined, represent an optimal construct of capability, vastly surpassing the sum of its parts.
SQREEM Data Ingestion Engine
Without being a multinational media, search, or social platform, SQREEM has developed techniques that allow for capturing the vast majority of online activity originating from any country. This is done independent of third party APIs or data providers and covers the widest imaginable spectrum of unstructured sources. It is also achieved in compliance with all privacy, copyright, and data-protection laws.
One of the biggest accomplishments in the company's seven year R&D evolution has been the capability to 'fuse' and combine virtually all data sources into a common topography. This allows for the detection/measurement of possible inter-dependence between dynamics or phenomena originating in vastly different areas on the information spectrum and is key in identifying true causality in whichever shape or form it may exist. You never know where a true cause may originate and what true effect it sets in motion, unless you have all relevant information and are able to measure all relationships it contains.
The SQREEM Engine
We can Ask the right Question to find the Right Answer Automatically
Measuring all possible relationships represents the apex of all challenges. While a combination of statistics, mathematics, and physics contains the theory and tools to design virtually any measurement, application requires enormous expertise and the design of unique measurements for unique relationships. This process is not scalable due to its complexity, even if the data is available. There is more causal uniqueness than grains of sand. Over half a decade of R&D, SQREEM has built an algorithmic platform which pragmatically applies the most universal theoretical foundation defining pattern and information from the ground up: Entropy. We have the capability to detect and measure pattern relationships and dependencies inside high entropy domains close to the theoretical limit and vastly beyond any comparable application in the world of data compression and encryption available today. In turn, SQREEM is able to systematically detect deterministic causality up to very high degrees of complexity using a singular approach.
Combining the ability to acquire vast volumes of data, fusing them into common topographies, and being able to detect complex, non-linear inter-dependencies introduces the final challenge we had to overcome. Real-world deployment requires massive amounts of computational resources. We had to design capability which allows dynamic scaling of data acquisition, fusion, and processing across devices, hardware-specs, operating systems, and physical locations. There is need for cost-efficient, seamless and elastically scalable virtualization that allows quick assembly of large scale computing power, using third-party data warehousing, VPNs, caching, and processing. Today SQREEM has its own platform combining internal resources with external service providers such as AWS, Digital Ocean, among numerous others, as well as secure, client-controlled domains, into one single dynamic deployment capability.