Large manufacturers rely on a vast network of thousands of suppliers, some of which may be located half a world away. When any one of those suppliers encounters a problem that impacts its production capabilities – be it natural disaster, labor strike, lawsuit, or government shutdown – the manufacturer must assess the situation and quickly decide on alternatives, before the entire supply chain is disrupted. Therefore, it pays for risk managers at large-scale manufacturers to closely monitor their supply chain partners and track potential supply chain risks.
However, it is nigh on impossible for a manager to identify relevant threats and mitigate risks in a timely manner across a widespread network. Even by combing through the information on the Internet, the largest body of knowledge available, a manager will conclude that the volume of data is too large and too widely distributed across various news sources, portals, and languages to be properly evaluated and understood.
But, must this really be the case?
Gaining knowledge in cyberspace, regardless of language
Frantisek Vrabel, CEO and founder of Czech technology startup Semantic Visions, asked himself this same question as he set out to build a Web-mining system that would comprehend the troves of information available on the Internet, regardless of language. Vrabel’s goal: to broaden people’s view on the world. “I wanted to create an all-encompassing system that provided a universal gateway to the world of knowledge, despite language,” says Vrabel.
Next page: SAP HANA provides faster data analysis
With a team of 30 top-notch engineers from 15 nations, Vrabel launched Semantic Visions in 2004. Having had successful startup experience in earlier ventures, Vrabel characterizes his life’s work as building on one central theme: knowledge discovery. “I grew up in the Middle East, in Iraq, and was fond of archeology. I was just excited about Mesopotamia and all the cultures around it,” he says. “I think this has formed my way of thinking and understanding that the world began a very long time ago and that it is increasingly interconnected.”
That notion of interconnected knowledge led Vrabel and his team to develop an advanced ontological algorithm that can read and extract meaning from unstructured data in 11 languages. Today, Semantic Visions has created the biggest semantic ontology in the world. It is the world leader in semantic analysis and open-source data processing. The company has built a semantic database comprising 35 terabytes of metadata, which is relied upon by not only manufacturers but also news organizations like CNN to track global events.
Semantic Visions works with SAP HANA for faster data analysis
Two years ago, however, Vrabel became frustrated with the speed of the processing technology underlying the data analysis for Semantic Visions’ applications. In meetings with his engineering team, he heard a common lament: “We don’t have enough computing power. It will take us weeks to get these answers and you’re asking us for a real-time tool.”
Vrabel began to search for a new type of technology that would enable a new class of services. “Finally, we concluded that SAP HANA would be the tool,” says Vrabel. “With traditional analysis and database tools, you simply cannot make a real-time application. From our perspective, SAP HANA enables us to make sense out of our huge, big data compilations.” Semantic Visions joined the SAP Startup Focus program and now uses SAP HANA as the basis to provide real-time data analysis in its applications. To learn more, watch the Semantic Visions video.
Vega is Semantic Visions’ real-time application, powered by SAP HANA, to detect and monitor supply chain disruptions. Each day, Vega reads 1 million articles in 11 languages from 160,000 sources and continually searches for supply chain threats. By combing the information available on the Web from sources such as news portals, company and government sites, and a selection of high-quality blogs, this application is able to gather content and conduct semantic analysis to render meaning.
Vega: supply chain risk detection in real time
Vega covers four supply chain scenarios: 1. direct threats, such as an industrial incident linked to a certain company; 2. threats in a certain geographical region, based on the knowledge that some suppliers are located in a specific geographical location; 3. sudden increases in the number of articles on a certain company, which might indicate that there is a problem; 4. sudden increases in negative sentiment articles.
Next page: How Vega assesses risk according to company
“The system is able to know basically about everything at any given moment,” explains Vrabel. “The challenge is to be able to discover the knowledge that is hidden inside. To do that, you have to have a powerful analytical tool. And you have to create a visual user interface that enables you to understand the complex results of the analysis.”
Vega categorizes the companies in the supply chain according to their level of risk exposure. It displays this information in a clear, graphical format. For those companies or threats that are marked red for high risk, the risk manager can then drill down to get more information. The application tracks this risk over time, so that risk manager can see if risk is diminishing or increasing. “What this application does is it gives you thousands of eyes and ears in addition to your own. It is a discovery detection tool,” says Vrabel. “You are not searching for something, because you can hardly search for the unknown. This system automatically alerts you that something may go wrong in your [supply chain].”
Find out more about how Semantic Visions’ Vega helps large manufacturers manage supplier risks.
SAP Startup Focus program
Semantic Visions is part of the SAP Startup Focus program, which supports startups in developing new applications on SAP HANA. SAP’s database platform streamlines analytics, planning and predictive and sentiment assessments to allow business to operate in real-time. The SAP Startup Focus program is an accelerator for Big Data, predictive, and real-time analytics startups. More than 650 startups are participating in the program to-date that have developed distinct SAP HANA-based use-cases from gaming to retail to finance and manufacturing.