The blog post details the process of creating an advanced agentic retrieval-augmented generation (RAG) system using LlamaIndex and Memgraph, building on previous setups by integrating graph algorithms such as PageRank. It describes setting up Memgraph as a graph store, defining function agents for data retrieval and arithmetic operations, and implementing a retriever agent to execute the PageRank algorithm and extract ranked nodes. The tutorial guides readers through designing an AgentWorkflow that combines retrieval and computation for automated query execution, ultimately enabling the system to dynamically retrieve and process graph-based data. By integrating PageRank into the agentic workflow, the post highlights the potential for enhanced data processing capabilities, suggesting further exploration of multi-agent systems and graph algorithms to build more intelligent and context-aware GenAI pipelines.