Our work is rooted in an enduring belief that thoughtful and systematic analyses of available data, coupled with a sound understanding of industry dynamics and trends give rise to informational advantage, which in turn precipitates measurable gains in the efficacy of business decisions.
With that in mind, we make it our mission to analyze emerging industry trends and key developments, continuously amass risk event data, leverage already-proven and emerging data analysis and visualization approaches and tools, all as means of developing valid and reliable mechanisms for estimating company-specific exposures to adverse developments.
Multisource and multivariate are the two key characteristics of our approach: The former reflects our emphasis on simultaneously analyzing source-dissimilar but outcome-related data types; the latter underscores our commitment to pursuing maximum amounts of knowledge that might be hidden in data.
In keeping with that mindset, our data analytic approach follows a simple, three step process: data -> information -> knowledge, which ultimately enables us to distil the commonly large volume of technical findings into a more manageable set of practically relevant decision drivers, offering clear and meaningful decision guidance.
How it works
To gain meaningfully deep insights into drivers of observed outcomes it is typically necessary to combine multiple, dissimilar sources of data. It is a complex task, especially if it is to encompass both structured (e.g., numeric) and unstructured (e.g., text) data sources - as captured by the high level summary graphic:
Text & Numeric Data
It has been estimated that more than 95% of all data is text, yet it is the numeric data that is the focus of about 95% of business analytical endeavours. Although the nuanced nature of human communications makes machine processing of text challenging, it is nonetheless feasible to search for, extract, and effect-code otherwise not available insights hidden in text data. When combined with numeric data, text-sourced metrics can significantly increase predictive and explanatory power of analyses.
We view analytic findings as the beginning, not the end of the data mining process - although we are cognizant of the value of outcome focused reporting, our ultimate goal is to turn generic informational outcomes into unique insights. While basic reporting can yield worthwhile 'what-is' information, it is the more involved analytics that offer decision uncertainty reducing 'what will be' and 'why' knowledge.
We believe that to be able to maximize the informational value of data, it is necessary to combine origin- and content-dissimilar data into a single informational reservoir. Joining together of multiple data files, often sourced from very unlike sources, is usually a multi-stage process, hinging on the identification of networked pairwise-communalities among individual data files.
We believe that the ultimate test of the value of data analysis-derived outcomes is the degree to which the resultant knowledge 1. leads to informational advantage, and 2. reduces the uncertainty associated with making strategic and tactical decisions. Our data analytic outcomes emphasize support competitively advantageous decision making by leveraging objective insights hidden in data.
Ongoing & Standardized
In order to be the source of competitively advantageous insights, analyses of data should be an ongoing, learning-focused process, built around distinct stages and generating cross-time and cross-type comparable outcomes.