The workshop on building knowledge graphs, led by Dan Shalev, explored practical implementations using VCPedia and Fractal KG, focusing on technical decisions, implementation patterns, and production considerations. Key topics included the use of large language models (LLMs) for automated entity extraction and relationship mapping, optimizing graph construction, and ontology-driven accuracy. The discussion highlighted challenges like entity resolution, memory optimization, and schema flexibility, with case studies illustrating the importance of deciding whether information should be modeled as nodes or attributes based on memory efficiency, traversal requirements, and query patterns. Additionally, the workshop addressed the scalability of ontologies with data updates, the benefits of storing information as graphs for context retrieval, and the need for more robust ontology enforcement mechanisms in schemaless systems like FalkorDB. The session concluded with insights on domain consolidation in knowledge graphs and strategies for improving attribute extraction accuracy using structured output methodologies and hierarchical context injection.