Solving Problems with Big Data Workflows: Flexibility vs. Reproducibility
Kineviz’ browser-based visual analytics platform—GraphXR—addresses a serious problem in conventional big data analytic workflows: lack of visibility during data provisioning, modeling, and analysis discourages exploration and discovery of new, unanticipated value in data products. Analytic workflows have typically been conceived of as straight-line, waterfall processes: an organization asks questions, gathers data, builds a data model based on expected patterns and current questions, conducts analysis, and finally visualizes the analytic results.
But it’s increasingly clear that new, unforeseen questions and valuable insights can arise at every step. The opportunity, and often need, to adjust project focus makes it ever more important to support flexible, iterative workflows that provide actionable insight along the way. Without this flexibility, data models can quickly become irrelevant, and adapting them for new scenarios becomes very cost- and technology-intensive for the organization.
Visual Analytics as a Solution
GraphXR is developed on the principle that the question should shape the data model, not the other way around. Hence, GraphXR’s in-memory analytics offer unprecedented flexibility to conduct refined and customized analyses without heavy lifting on the data management side, and without affecting the underlying data. Using GraphXR, analysts can dig deeper into data warehouses or lakes to easily map and find connections between high-dimensional and varied data sets as needed.
To facilitate such a streamlined approach, GraphXR supports accessing data sources through query of structured or unstructured data stores, or simply drag and drop. This empowers organizations to gather and fuse data from anywhere. During analysis, GraphXR brings the best of both worlds— network and statistical analysis—to reveal needle-in-haystack discoveries that evade traditional analytic workflows.
Who Can Benefit and How
While GraphXR’s enterprise edition can be used as an off-the-shelf product, Kineviz also provides consulting services to help clients implement modern iterative in-house analytics. Kineviz customers typically have high security requirements, needing on-prem and even air-gapped deployments, with custom pipelines and workflows.
Recent successes include a law enforcement client monitoring activity of a terrorist network on a social media platform. With massive data at hand, they struggled to obtain clear actionable results that could be used for operational planning. With GraphXR, the law enforcement agency was able to deplatform the terrorist entity, curbing its recruiting and propaganda. Kineviz and GraphXR have delivered success stories such as this in many other sectors as well, including biotech, cybersecurity, manufacturing, and logistics. Today, GraphXR is the cornerstone of many groundbreaking projects, including the National Institute of Allergies and Infectious Diseases’ research on COVID-19 spike proteins.
Our goal is to put actionable information in the hands of people who need it and understand it best
By going beyond trendspotting to actually identify patterns, GraphXR is enabling organizations to discover opportunities, obstacles, and potential black swans, and ultimately mine the gold in their data.