Shortest Path, Instantly: How to Navigate Millions of Nodes Without the Lag

​Roi Lipman, CTO & Co-founder and Gabriel Gomez, Core Software Devoloper break down how FalkorDB finds shortest paths at scale.

​Finding shortest paths in a graph seems simple, until the graph gets large enough that it isn't. Even a small, ordered search space produces millions of possible routes between two points. At real scale, most graph databases fall back on the same handful of algorithms, and all of them get slower as the two points get farther apart.

​We took a different approach. FalkorDB is the first graph database to ship Customizable Contraction Hierarchies (CCH) as a built-in algorithm, alongside highly optimized variations of Dijkstra and A*. In this session, we'll walk through what that actually means in practice.

​What we'll cover:

• ​Why shortest path search gets expensive as a graph grows, and why that's true even for algorithms considered "fast"

• ​An easy to understand explanation for where Dijkstra, A*, and CCH make the most sense

• ​What we found in our own testing: query cost that stays roughly flat as trips get longer, instead of climbing with distance

• ​Real use cases where this matters: logistics and delivery routing, network and infrastructure path planning, and large graph traversal generally

​Who should attend: engineers and architects working with large graphs where path-finding performance is a bottleneck, and anyone evaluating graph databases for a workload with this kind of query pattern.