EnterpriseWeb founder and CEO, Dave Duggal was a guest on the Adorsys podcast to explore EnterpriseWeb’s innovative use of hypergraphs (Podcast Link).
The hosts, Tim Biedenkapp, Inga Glotzbach, and David Roldán Martínez led a lively discussion, which endeavored to make the highly-relevant, but deeply technical concepts accessible to a wider audience. As Dave relates in their conversation, Hypergraphs are an advanced form of graph that supports multi-dimensional and conditional relationships. Historically, this approach has been limited to academic research, but EnterpriseWeb provides a practical no-code solution that enables architects and engineers to leverage the power of hypergraphs while abstracting away the complexity of modeling and processing.
Demystifying Graphs
Graphs provide a way to model domains based on relationships, just as social graphs are understood to manage relationships between people. The advantage of graph relationships is that they can be nested (e.g., direct or 1st degree, 2nd degree, 3rd degree, etc.) to capture interrelationships and interdependencies. New relationships can be added and old relationships can be dropped, so graphs can evolve over time. Logical relationships are naturally machine-readable so humans and agents can traverse and query the graph.
Conceptually, a business is a graph of relationships between people, information and capabilities bound by events and policies. Likewise, distributed systems can be viewed as relationships between the collection of nodes in a network, and services represent the set of relationships between functions and service endpoints.
Graphs can be used to design ontologies. Ontologies model the relationships between high-level concepts, types, policies of a domain, abstracted from any specific implementation or use case. Domain entities are mapped to the Ontology, which makes it a common language for reasoning and interoperability across the domain. Ontologies are highly-relevant to advancing Generative AI and Agentic automation, which both depend on external context to optimize their outputs and actions.
Enter Hypergraphs
Of course, relationships can be complex, which is where hypergraphs shine as they naturally support 1:many relationships, which can be conditional – based on business logic. Hypergraphs allow for a more dynamic and contextual representation.
EnterpriseWeb extends the use of hypergraphs to complex, real-world business operations. To do this, the company leverages hypergraph dimensions to model business logic, functions, workflows, governance policies and temporal history all in a single graph, alongside the ontology (i.e., domain semantics, metadata and entities). The hypergraph is a rich digital representation of the business. Whereas conventional graph technologies like Semantic Web and Labelled Property Graphs are all about data (data analytics, recommendations and science), EnterpriseWeb’s use of hypergraphs is focused on enabling real-time, enterprise-grade, contextual automation.
The company’s platform allows users to model their domains, use cases and objects in a fully declarative manner at design-time, with the platform’s agents translating user-intent and handling implementation details. At run-time, platform agents leverage the hypergraph to interpret events and requests, evaluate related conditions, contextualize decisions and optimize actions. EnterpriseWeb agents can compose and configure tasks into complex workflows and orchestrate distributed solution elements, including LLMs. Agents dynamically enforce governance and business logic on-the-fly in the same pipelines for fine-grained policy enforcement.
EnterpriseWeb bridges knowledge engineering and programming languages for ontology-driven, agent-based automation.
Related Links:
Sorry Telcos, graph databases are no way to run a business
Semantic Web and Labeled Property Graphs have their use-cases, but complex, real-time business operations is not one of them
Agentic Automation: Knowledge is Power
Agent-based automation must bridge AI and structured knowledge to improve inferences and optimize outcomes
The Telecom Ontology: The Missing Links for Level 4 Autonomous Networking
To unleash the value of network data and enable dynamic configuration and continuous optimization, Telcos need a unified and harmonized network model. A standards-based Telecom Ontology provides the missing links, a semantic layer, to connect, query, simulate and manage operations!
Lowering the cost of context for real-time, data-driven systems
Assessing the technical and organizational roadblocks to intelligent, end-to-end automation
