Zachary Blumenfeld distinguishes execution graphs, context graphsA knowledge graph represents entities and their relationships in a structured form that can be queried and connected. and procedural graphs as ways to represent agent workflows, decision traces and reusable instructions. He argues that storing these structures alongside data can support portabilityInteroperability is the ability of different systems to exchange information and use it with compatible meaning and behavior. and governanceGovernance defines responsibilities, decision rights and controls for how a system or organization is operated. across agent runtimes.
In a Neo4j demonstration, Blumenfeld describes compiling a loan approval procedure into graph-shaped steps, scripts and routing decisionsAgent orchestration coordinates AI agents, tools, people, tasks, state, and control flow so a larger workflow reaches a verified outcome.. He shows how retained execution tracesAn execution trace records the sequence of operations and results produced while a system carries out a task. can help inspect branches and assess a workflow over time.
The speaker presents an informal demonstration of reduced execution time and cost, while noting that it is not a formal research evaluation.
Watch on YouTube
