Telco-grade Generative AI for Intent-based Orchestration, Part 1

Dave Duggal, founder and CEO, EnterpriseWeb 

Generative AI has captured the public’s imagination with its human-centered interfaces (text and voice) and its advanced conversational capabilities, which make the power of artificial intelligence accessible to non-technical people. There is a race already underway to extend generative AI use-cases and capabilities to virtually all aspects of human endeavor, including the large market for business productivity and automation solutions.  

While generative AI has already spawned much experimentation, most demonstrations are focused on the technology’s ability to generate outputs – text, music, art, and even code templates – as guided by human prompts. These outputs, while impressive in their own right, are generally understood to be first drafts, often with varying degree of errors and omissions, such as an entry-level assistant might provide. Early generative AI automation prototypes tend to demonstrate simple if-this-then-that (IFTTT) style capabilities without reference to security, governance and compliance concerns. Complex enterprise and industrial systems have a far higher standard to safely automate processes with complex use-case requirements, organizational policies and industry regulations, which require deep domain-specific knowledge and rules.  

EnterpriseWeb’s no-code platform was designed specifically to allow software engineers to rapidly design, deploy and manage highly-dynamic, data-driven applications. It features a graph knowledge base, which provides domain context, configuration and control for stateless middleware functions providing intelligent back-end services for developers. Now, with the advent of generative AI the company can expose its declarative interfaces to developers and AI for collaborative design, deployment and management of next generation business and infrastructure applications. EnterpriseWeb’s integrated best-of-breed solution with KX advances the state-of-the-art by providing the necessary domain-specific context and constraints to enable telco-grade generative AI for network service orchestration.  

Our first Generative AI demo, now available on YouTube, presents a developer interactively composing a network service in an informal conversation with generative AI. It features Microsoft Corporation’s Jarvis project, which provides a natural language programming interface and uses a foundational model from OpenAI. KX’s vector-native, time-series database, acts as an intermediary between generative AI and EnterpriseWeb’s automation capabilities. This purposeful design provides an architectural separation-of-concerns between the two technologies, which serves as an important security and IP boundary. As the intermediary, KX translates between the developer’s unstructured natural language requests and the structured, rules-based logic of EnterpriseWeb’s graph domain model. EnterpriseWeb interprets the queries and commands using the graph to identify relevant relationships to known business concepts, types and policies. It allows the developer to express their intent in an informal short-hand, with the platform runtime handling the implementation complexity.  

Once the developer has configured the network service, they can place an order directly through EnterpriseWeb or a front-end like Microsoft’s PowerApps. In either case, EnterpriseWeb provides the back-end to automate the deployment and configuration of the service. The platform’s run-time takes responsibility for enforcing system controls, IT governance and business compliance, which represent strong rules-based guardrails for safe systems behavior. Post-instantiation, KX observes the network service and reports events to EnterpriseWeb for closed-loop autonomic management (i.e., self-scaling, self-healing, self-optimizing). 

Together, EnterpriseWeb and KX are providing an integrated generative AI automation solution for next generation AI-powered Telecom operations, which is now available for commercial trials. The solution itself is generic; Telecom is the first domain being demonstrated. EnterpriseWeb and KX are working with partners to develop and deliver generative AI automation across industry verticals.