Beyond Joins – Supply Chain Optimierung & Next-Gen AI mit Graphen

16:00 - 16:30

Doors Open

16:30 - 17:00

Intro: The Paradigm Shift

From the relational model to graphs: why deep SQL joins become slow and hard to manage in connected supply chains, how the “join problem” emerges in questions like “who supplies the supplier of my supplier,” and why index-free adjacency lets graph traversals run in constant or linear time instead of exploding in complexity. A direct comparison of modeling a bill of materials in SQL versus a graph makes the core difference clear: in graphs, relationships are first-class citizens; in SQL, they are only metadata.

17:00 - 17:45

Deep Dive: Supply Chain Management (SCM) with Graphs

Practical application: using graphs to bring transparency, compliance, and risk control to supply chains that function as networks, not linear chains. Examples include end-to-end traceability across bills of materials, simpler and faster queries compared to recursive SQL, and impact analyses that show which products are affected by disruptions. Graphs also expose single points of failure and enable route optimization based on cost, CO₂, and time.

17:45 - 18:00

Coffee Break

18:00 - 18:45

Innovation: AI & GraphRAG

From hype to application: how graphs make AI smarter by adding structure and context to RAG. Vector-only search often misses relationships when answers depend on linking multiple facts. GraphRAG combines unstructured text with structured knowledge graphs, using the graph as memory and fact-checker to reduce hallucinations. In a supply chain scenario, vectors find documents about a strike, while the graph connects logistics providers, parts, products, and revenue impact—allowing the LLM to deliver a grounded, business-relevant answer.

18:45 - 19:30

Networking - Dinner