Graph Computing - the Next Big Wave in AI

By Susan Hu


Whether we are aware or not, Artificial Intelligence (AI) has already been powering aspects of our work and day-to-day life over the last few years. It seems to be only a matter of time before we will live in an AI-powered new world. As AI hype rises, we can't help but wonder what the next big breakthrough will be to push us towards the next wave of AI applications. Graph computing is at the top of the watchlist.

Today, we are going to explore what Graph Computing is, why it's the next big wave in AI and how it can help you and your business thrive in the next-gen AI era.

What is Graph Computing (and Why Does it Matter)?

The human brain is a giant graph of 100 billion nodes and 700 trillion edges. Our brains are extremely good at cross referencing data from different senses and connecting the dots. In other words, graphs are the natural way to store and process unstructured data in the real world, aka how human brains think.

For example, when solving complicated problems, human intelligence naturally observes different components of a situation and, based on previous experiences (knowledge base), it connects dots and comes up with solutions.

To truly achieve next-gen AI, which mimics and exceeds full human brain functions, it is essential to gather, store and process information in a similar way that does not look at information as it's isolated, but rather consider the complex connections and relationships between each piece.

When it comes to complex real-life problems such as transactions, relationships, events, medical assessments, human behavior, psychology, etc., it is fundamental to gather, store and process information in a connected way. Traditionally, when we process data, it has always been the two-dimensional method, making it impossible for computers and machines to solve any real-life problem beyond simple calculations and automation. That's why we believe to achieve the true AI, Graph Computing is an essential foundation, and it will be the next big wave to pushes AI to the next era.

Alt Text An example of a graph, showcasing different entities and their complex relations.

Applications of Graph Database and Analytics

Graph Computing and its applications is already a big part of our day to day life – without most of us knowing it. For example, when you turn on Netflix, it often gives you recommendations of what to watch next. What you may not know is, Netflix has its algorithm automatically calibrated to your interests, viewing history and based on its knowledge graph database, made recommendations for you! Similarly, when you go online shopping, you often see "items you may also like" as you are ready to check out. Guess what, this is all Graph Computing at work.

Of course, Graph Computing can, and is already being used in other more complicated problems across industries.

Here are a few use cases how Graphen utilizes Graph Computing as a foundation to store and process data and build our true AI foundation and solutions to perceive, process, strategize and solve problems.


Using Graph Computing as a foundation, Graphen developed and deployed an AI powered self-defense cybersecurity system at one of the largest banks in the world. The system monitors and analyzes user behavior to detect insider threats within the organization. Built upon SIEM (Security Information and Event Management), the system analyzes user, device, application and network event data to create and update entity behavior models at individual and group levels in real time. It then compares current behavior models against historical and group models to flag deviations and detect outliers, send an alert for insider threat within the organization.

Anti-financial crime

With relationships in consideration, Graph Computing makes it possible to work with huge volume of datasets, detect and prevent sophisticated financial crime in real time. Graphen uses Graph to process and analyze financial transactions and other activities in real-time. Comparing to conventional fraud detections, Graph Computing reduces false positives, detects Unknown-Unknowns. Graphen has built customized solutions based on Graph Computing for Anti-Money Laundering (AML), Anti-Fraud Detection, Trade Finance Due Diligence, Regulation Interpretation and Audit.

Market intelligence

Build upon Graph Computing, Graphen Market Intelligence solution helps financial institutions and analysts make informed investment decisions. The solution automatically collects public information, forming knowledge graphs and database. It learns from financial statements, announcement, news and analyst reports, construct knowledge graph, link cause and effects. Real-time news with relevance rankings according to individual portfolios and watchlists were automatically extracted. Graphen then use its advanced Machine Learning to aggregate news impact score on stock price.

Precision medicine

The future of precision medicine relies on a good understanding of the causality. Built upon Graph pathway analysis, Graphen's health AI can accurately predict the progress an individual's disease development and suggest potential treatment precisely based on that individual's gene sequencing and health background.

Personalized healthcare and prediction

Built upon graph pathway analysis and Graphen's medical intelligence database, it is now possible to accurately and proactively offer personalized suggestions for potential disease prevention including personalized heath risk analysis, prevention recommendations, and recommendations of relevant health news information.

What's the Future of Graph and AI

From e-commerce to online streaming, finance to healthcare, we can see Graph Computing is already a part of our everyday life. Understanding what Graph Computing is and how it stores, processes and analyzes data, we believe, there will only be more and more applications in our day-to-day life in the next couple of years.

As businesses recognize the importance of the interconnected way data should be processed, they will start to leverage Graph Computing to process a much bigger volume of data and solve more complex problems. Graph Computing will mature to become easier to adopt, plug-and-play kind of adaption.

With the growth of Graph Computing as a foundation, Machine Learning, especially autonomous learning will be able to break through the current bottleneck limited by both the quality and quantity of data. When we have a superior knowledge database, it might not be long till we see the true next-gen AI achieve and exceed human intelligence.

If you want to learn more about Graph Computing or how it can be a strong foundation for your AI adaption and transformation, reach out to me at susanhu@graphen.ai.