What we did

Analytics

Client

Global Technology Firm

Completion

5.8.2023

Utilizing A Knowledge Graph to Manage Business Risk and Complexity

Global Technology Firm Implements Neo4j
black arrow pointing down

Utilizing Graph Databases To Manage Complexity Among Users, Authorization, and Intellectual Property

Schedule a call

Our customer, a premier, global research organization, produces large quantities of intellectual property in the form of whitepapers, webinars, and articles.  They also provide high-value, analyst-based, advisory services to organizations requiring a combination of business and information technology insights. The customer wanted the ability to exercise more fine-grained authorization control and faster content access, both to ensure a positive customer experience and reduce potential revenue leakage.

The customer required a solution to manage the ever-expanding number of relationships among the forms of content produced and their growing customer base with fine-grained content access requirements.  A fast and efficient means of determining which users and organizations were authorized to access what content was required.  Managing relationships between content and user authorization at run time is complex and if not done well will be the source of a poor user experience. Worse, it can be the source of the revenue leakage attempting to be addressed in the form of unpaid access to content and services.

Skylytics architected a solution that combed microservices, data streaming, and a Neo4j Graph database to deliver a reliable, performant, and scalable solution to solve this challenge.  Customer Authorization, content metadata, detail, and Taxonomy are streamed in real-time from multiple upstream sources, the graph is kept in sync with the upstream systems to allow for successful management of the complex relationship that exists between users and content.  

The Neo4j database enabled the implementation of a comprehensive Knowledge Graph to model the content, users, and authorization.  Knowledge Graphs have emerged as a powerful tool for businesses to remain competitive in rapidly evolving industries. Organizing data using knowledge graphs allows users to efficiently answer complex questions and gain insights from interconnected data sources where traditional relational databases fall short.

BI and Data Science Benefits Provided by the Knowledge Graph

The Neo4j Knowledge Graph provides the customer with additional benefits as well. The Graph enabled them to move away from hardcoded access rules and towards a more modular and flexible approach using Cypher queries. A reduction in the number of connections between services previously needed to complete the authorization task was also achieved.  They can now also answer questions that were previously difficult or near impossible to answer with traditional relational databases due to the number of many-to-many relationships that exist.  

Neo4j has also helped facilitate Graph Data Science using embedded features in Neo4j and as well as enhancing their Business In.  The company is now able to address potential customer churn, potential revenue leakage, basket analysis, and clustering to expose actual customer segmentation. These capabilities enable their business to proactively address customer concerns, ethically target new business, minimize potential revenue loss, and deliver a more seamless user experience.

Knowledge graphs have demonstrated their potential to revolutionize traditional data management practices and optimize business operations, particularly in ecosystems with multiple data silos and complex relationships. The implementation of a knowledge graph fosters a more data-driven culture and enhances the capabilities of Business Intelligence while delivering valuable benefits such as better customer retention, maximized revenue opportunities, and improved user access to relevant information.

“Knowledge Graphs enable performant, scalable, real-time insight across the business landscape. Graph databases organize and expose complex relationships among vast numbers of business entities that may not be obvious or known to an organization. We’ve utilized graph at the core of multiple customer solutions including identifying component sourcing risk in complex supply chains, and proximity, process, and resource risk in IoT-based asset tracking solutions.” – Michael Hickey, Co-Founder Skylytics‍
Confidential
Global Technology Firm

Our approach

The client continues to enhance backend systems that serve content to users who have purchased it and are authorized to access it. There is also an effort to better understand the relationships among the producers of content, the consumers of the content, technology themes, and access patterns. Graph is core to the program.
View live
01

ID'd Relevant Entities & Relationships then Designed and Implemented a Label Property Graph

Defined the scope of the project and identified the entities and relationships that were to be included in the graph. Examined each source of truth to determine where the information about entities and relationships resided and how often this data changed. Created an unstructured schema which included identifiers, properties, and relationships for each node and edge, ensuring that the model would enable answering the competency questions defined in advance. Used the Label Property Graph (LPG) model in the Neo4j Graph Data Platform to store the knowledge graph.
02

Developed Microservices and used Apache Kafka for Data Synchronization

Created three services to perform specific tasks like communicating with third-party providers, keeping the graph in sync, and handling permission access rules. Leveraged Apache Kafka to keep the graph database synchronized with real-time changes.
03

Integrated with Business Intelligence Platform

Collaborated with the client's BI team, providing them with the right Cypher queries to extract data from the graph and integrate it with all the existing data in the ecosystem.
04

Visualized and Verified Data

Developed a knowledge graph with greater than 50k nodes and 2 million relationships, enabling the client to answer complex questions in real-time and integrate data with other business intelligence tools
Explore
Start the conversation

Let's start building your continuous intelligence solution

Schedule a call