Virtualization Technology News and Information
VMblog Expert Interview: Katana Graph Talks Impacts of Graph Technology and Analytics

interview katana graph 

VMblog reached out to Farshid Sabet, Chief Business Officer of Katana Graph to discuss the status quo of graph technology, and find out what steps both graph vendors and businesses can take to get the technology to scale.

VMblog:  Graph analytics and AI is a fast-emerging technology. What were people doing prior to graph analytics in business?

Farshid Sabet:  Prior to graph technology, organizations were using alternative forms of data analysis technologies that were more geared towards structured data. These technologies are less efficient because they are able to analyze only structured data, pose limits as to what size of datasets can be analyzed, and don't provide the deeper insights that comes with analyzing unstructured data - a type of data that cannot be easily categorized into rows and columns, such as audio, video, social media postings, and other units. Now, with 95% of businesses citing that managing unstructured data is a serious challenge, graph computing is quickly emerging as the emerging solution to help many organizations across varying industries understand and obtain insights from troves of complex data sets and better advance their business goals.

The technology evaluates trends between basic units of data structure, nodes, and their relationships, edges, to find patterns that data teams can derive further insights and next steps to improve processes. Graph technology's importance comes at a critical time as unstructured data makes up 80% and more of enterprise data, and is growing at the rate of 55% and 65% per year.

VMblog:  What are the chief challenges or barriers that businesses face when adopting graph technology?

Sabet:  The two main barriers that businesses face are the awareness of graph technology and the technology's own interoperability. Despite graph technology's incredible ability to help organizations understand huge sets of unstructured data, organizations are still in the dark about this emerging technology - including the insights and benefits it can bring. This challenge will naturally wane as graph technology inevitably proves its worth.

One example of graph's success includes its ability to mitigate cyberthreats. By examining large sets of data to track how networks of endpoint devices access each other and external points of contact, graph computing can detect any intrusive or alerting points of access from outside the trusted network or suspicious devices. Here, graph technology is already proving to be able to detect fraud, breaches of privacy, and other security vulnerabilities at breakneck speed.

The second barrier is graph technology's ability to interoperate at scale with other coding and data storage languages, such as Spark and other external libraries. Due to graph's relative novelty, there are still limitations on the size and type of data that can be processed. Fortunately, the ingenuity of graph technology is pushing against its own limits and integrating itself with other data languages to expand capabilities and provide better results. For example, in my experience, an identity management company was having trouble fully understanding their data to scale. The graph platform was able to operate with Spark, a data processing alternative, but with many additional machines and excessive costs. Once the company shifted over to a graph technology platform that was well integrated with Spark, the organization had better agility and lower costs with data transfer and analysis.

VMblog:  How will the rise of graph technology impact or benefit data scientists? Given the shortage of data scientists, how can graph analytics/AI make it easier for data scientists - and other roles - to be more efficient or effective?

Sabet:  Data scientists play a crucial role due to the growing amount of unstructured data and the vast potential of knowledge that it shows. With graph technology, data scientists will play a more important role within their organizations, as graphs are a better tool to derive data analysis and provide insightful recommendations.

Graph technology will not only improve the work processes and capabilities of data scientists on average, but also lessen their workloads. With data - and the insights derived from that data - being arguably an organization's most precious resource, the data scientist role continues to increasingly be a critical business-focused role that is enjoying a more influential presence.

VMblog:  Can you provide me with specific examples where you're seeing graph analytics/AI deliver real-world benefits to organizations?

Sabet:  Another sector that graph technology is empowering with deeper data insights is the pharmaceutical industry with drug discovery. In this context, graph examines data sets full of varieties of medicinal drugs, treatment plans, certain chemicals, and patient characteristics to analyze what kind of response would the patient's injury or illness have. Pharmaceutical data scientists then evaluate such relationships to provide better and more holistic predictions on what drug would work the best for a certain treatment for a particular patient. As drug discovery teams look to avoid the long and arduous trials of wet-lab experiments, graph technology can save resources and time, saving lives faster.

VMblog:  Are there any other up-and-coming use cases that you are seeing?

Sabet:  One emerging area for graph is its application across distributed computing systems and the hybrid cloud. As remote/hybrid work continues, both people, devices, and data are becoming strewn across geographically. However, data is costly to transport, and it's essential that data is organized into separate silos to help isolate and secure data. Graph technology is able to analyze this data across silos in unified scope and provide data insights for the new landscape of distributed computing systems.

VMblog:  Are there any common misconceptions or myths about graph technology?

Sabet:  One downside to graph technology is that it is extremely demanding in terms of processing and data management requirements. Some organizations look to this as a reason to avoid integrating graph technology to their arsenal of data analysis tools. However, implementing any new way of organizing and analyzing data can seem overwhelming at first, but in the long run, graph technology provides deep data insights for data scientists to study and recommend next steps that will increase revenue, increase efficiency, and improve organizational processes.

VMblog:  How do you see the graph industry changing in the next year? In the next five years?

Sabet:  As graph continues to integrate with artificial intelligence, machine learning, and distributed cloud systems, not only will the technology improve but the application of graphs will expand across virtually every industry - from financial services, pharmaceutical, technology and many more. As a fast-emerging innovation, graph technology is just scratching the surface - I expect that the technology will play a more influential role in diverse use cases - from mitigating cybersecurity attacks, accelerating drug discovery, improving the customer experience, and many other applications that the graph industry is yet to realize. Going back to the first challenge of graph's lack of awareness within the enterprise community, this expanding role of graph technology will continue to spur its own growth.


Published Tuesday, December 21, 2021 11:30 AM by David Marshall
Filed under: ,
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<December 2021>