Agent Builder and Framework — Introduction
K2view’s GenAI Data Fusion framework, referred to as AI Fusion, provides a robust foundation for developing AI-powered agents that seamlessly integrate large language models (LLMs) with enterprise data and operational processes. It is built on top of Fabric — the K2view Data Product platform — and connects structured data from unstructured sources such as corporate knowledge bases and procedures.
This framework enables AI agents to utilize context, plan actions, and orchestrate complex workflows to manage conversations and deliver accurate, data-driven answers.
The agent framework consists of Agentic workflows that are built as Broadway flows in the project's implementation, improving reliability and flexibility. This approach gives the organization and its implementors control over the process and greater flexibility. It allows them to combine Broadway-based logic, code parts, and LLM calls, with full debugging capabilities.
The following articles further explain the core components, architecture, agentic workflows and recommended development patterns for implementing AI agents.
Agent Builder and Framework — Introduction
K2view’s GenAI Data Fusion framework, referred to as AI Fusion, provides a robust foundation for developing AI-powered agents that seamlessly integrate large language models (LLMs) with enterprise data and operational processes. It is built on top of Fabric — the K2view Data Product platform — and connects structured data from unstructured sources such as corporate knowledge bases and procedures.
This framework enables AI agents to utilize context, plan actions, and orchestrate complex workflows to manage conversations and deliver accurate, data-driven answers.
The agent framework consists of Agentic workflows that are built as Broadway flows in the project's implementation, improving reliability and flexibility. This approach gives the organization and its implementors control over the process and greater flexibility. It allows them to combine Broadway-based logic, code parts, and LLM calls, with full debugging capabilities.
The following articles further explain the core components, architecture, agentic workflows and recommended development patterns for implementing AI agents.
