Product Name: IBM Graph (formerly offered as part of IBM Cloud), Db2 Graph (part of IBM Db2)
Company Name: IBM
URL: https://www.ibm.com/products/db2 (Db2 Graph)
Entry Year: 2016 (IBM Graph launched on IBM Cloud, Db2 Graph introduced in 2020)

Graph DB Revenue (or Market Share):
IBM does not disclose specific revenue or market share for its graph database solutions, but it is part of the larger IBM data and AI portfolio. As IBM Db2 is a widely used database solution in enterprise settings, Db2 Graph extends its graph capabilities to Db2 customers.

Number of Employees in the Graph DB Division:
Exact figures for the graph division are not disclosed, but IBM’s data and AI teams are large, with significant resources dedicated to database and AI technologies.

Major Users:
IBM Graph and Db2 Graph are used by organizations in various industries, including banking, healthcare, government, and telecommunications. Major users of IBM’s broader database solutions include companies like Citibank, Kroger, American Airlines, and the US Department of Energy.

Key Application Areas:
IBM Graph and Db2 Graph are applied in fraud detection, network analysis, social media analysis, recommendation engines, and enterprise knowledge graph development. Db2 Graph is commonly used in financial services and telecommunications for analyzing large-scale, interconnected data.

Product Overview:
IBM Db2 Graph is an extension of the Db2 database that brings property graph capabilities to the well-established Db2 platform. It allows users to run graph queries on data stored in Db2 without migrating it to a separate graph database. IBM Graph (on IBM Cloud) was a fully managed graph database built on Apache TinkerPop, but it is no longer actively promoted as a standalone service. Db2 Graph focuses on integrating graph functionality into enterprise relational databases.

Data Compatibility:
Db2 Graph supports querying data stored in Db2 databases and allows the combination of relational and graph data. It uses SQL for relational data and Gremlin for graph traversal. Db2 Graph can ingest data from relational, NoSQL, and other enterprise data sources through Db2’s powerful data integration tools.

Knowledge Graph Implementation:
Db2 Graph can be used to implement knowledge graphs using its property graph model. It allows users to explore relationships and complex networks within their relational data, transforming enterprise data into actionable insights by combining traditional SQL queries with graph analytics.

Query Method:
Db2 Graph uses Gremlin as its primary graph traversal language, allowing users to perform complex graph queries alongside traditional SQL queries within the same Db2 instance. This hybrid approach makes it easier to apply graph analytics to existing data without requiring new data infrastructure.

Natural Language Queries:
IBM does not natively support natural language queries within Db2 Graph, but it offers integration with IBM Watson services like IBM Watson Natural Language Understanding and IBM Watson Assistant. These can be used to interpret natural language inputs and convert them into SQL or Gremlin queries for graph and relational data analysis.

Native Machine Learning:
IBM Db2 integrates with IBM Watson Machine Learning and IBM Cloud Pak for Data to support machine learning workflows. While Db2 Graph does not have built-in graph-based machine learning, users can export graph features and apply them to machine learning models using these tools.

Support for Traditional Machine Learning:
Db2 Graph can be integrated with IBM Watson Machine Learning and other AI tools to extract graph features for traditional machine learning models. This allows users to perform tasks such as classification, clustering, and prediction on graph-enriched data.

Support for LLMs:
IBM has been exploring the use of large language models (LLMs) through its IBM Watson AI platform. While Db2 Graph itself does not natively support LLMs, it can be integrated with Watson services to enhance text-based analytics and knowledge graph generation by using LLMs to extract entities and relationships from unstructured data.

Support for RAG (Retrieval-Augmented Generation):
Db2 Graph can be part of RAG models by acting as a data source for fast retrieval of structured graph data. This real-time, in-database retrieval enhances large language models’ text generation by providing relevant, contextual graph-based information for more accurate outputs.

Other Notable Features:

• Hybrid SQL and Graph Queries: Db2 Graph allows users to perform SQL and graph queries side by side, combining relational and graph data analytics within the same database system.
• Scalability and High Availability: As part of Db2, Db2 Graph benefits from Db2’s scalability, security, and high availability features, including multi-cloud deployment and disaster recovery options.
• Integration with IBM Cloud Pak for Data: Db2 Graph is fully integrated with IBM Cloud Pak for Data, a comprehensive data and AI platform that offers advanced analytics, machine learning, and data management tools.
• Security and Compliance: Db2 Graph inherits Db2’s enterprise-grade security, including role-based access control, encryption, and compliance with industry standards like GDPR and HIPAA.
• Graph Analytics: Db2 Graph offers powerful graph analytics capabilities for analyzing complex, connected datasets, particularly useful for large enterprise environments.