Neo4j Live: Inside StrangerGraphs – Predicting Season 5 with Graph Intelligence
Chapters 11
- Play from Welcome & Intro
- Play from Graphs as digital twins & learning through play
- Play from Graph data science as “alchemy” (communities & interpretation)
- Play from Origin of Stranger Graphs & prediction idea
- Play from Live demo: character agents & graph-powered predictions
- Play from Exploring the graph: prediction communities & accuracy
- Play from Pipeline overview: scraping Reddit & fandom data
- Play from Chunking, classification & managing LLM costs
- Play from Similarity graphs, communities & prediction hubs
- Play from From predictions to outcomes: matching plot points
- Play from Q&A
Join us as we break down StrangerGraphs, a prediction graph built from Reddit fan theories, Neo4j AuraDB, GPT-5 analysis, and GraphRAG-powered agents to explore what the Stranger Things community got right in past seasons - and what they might reveal about Season 5.
Key Highlights
- Reddit prediction mining + GPT-5 accuracy scoring
- Leiden clustering to find high-signal predictor communities
- Season 5 predictions extracted from accuracy-based hubs
- AuraDB + GraphRAG powering character-aware AI agents
Guest: Henry Collie
Resources mentioned
GenAI & GraphRAGAuraDB
More from Neo4j Live
- Watched Duration: 1:04:55
- Watched Duration: 53:27
- Watched Duration: 1:05:33
- Watched Duration: 57:08
- Watched Duration: 41:43
- Watched Duration: 1:08:27
- Watched Duration: 55:39
- Watched Duration: 1:05:41
- Watched Duration: 1:02:18
- Watched Duration: 59:06
- Watched Duration: 58:43
- Watched Duration: 59:07
- Watched Duration: 1:03:26
- Watched Duration: 2:22
- Watched Duration: 57:29
- Watched Duration: 29:59
- Watched Duration: 20:53
- Watched Duration: 3:03
- Watched Duration: 13:22
- Watched Duration: 7:35
