The system described takes unstructured text documents and extracts knowledge in the form of Subject-Predicate-Object (SPO) triplets. It uses a large language model (LLM) of your choice to identify entities and their relationships, visualizing these connections as an interactive knowledge graph. The project supports various OpenAI-compatible API endpoints, allowing for flexibility in implementation. Users need Python 3.11 or higher and must install required dependencies via `pip` to get started. Key features include automatic text chunking, which splits large documents into smaller, manageable sections, making them suitable for processing by the LLM. The knowledge extraction process identifies entities and relationships, followed by entity standardization to ensure consistent naming across the text. The system also infers additional relationships between disconnected parts of the graph, enhancing the overall coherence of the knowledge representation. The interactive visualization helps users explore the relationships between entities easily. During a demo using an Industrial Revolution text, the system processed 13 chunks, extracting 216 triples and standardizing 201 unique entities to 181. It then inferred new relationships, resulting in a final knowledge graph containing 564 triples. The output includes a saved HTML file for visualization, revealing 161 unique nodes and 564 edges, with 355 inferred relationships. This comprehensive approach minimizes fragmentation in the knowledge graph and provides a clearer picture of the document's content.