~/talk

Modern Graph Databases

Talkat NDC Oslo 2026 · 2026-09-14..09-18· [databases, machine-learning]
slides:none on file

Current graph landscape gets specific: Neo4j remains the dependable OLTP choice; Kuzu is bought by Apple and the open fork Ladybug (Kuzu fork) surfaces as an embedded analytical graph DB used inside Python/Jupyter to import CSVs, run Cypher, and quickly find cycles (demo finds Bob↔Carol and longer laundering cycles). PuppyGraph acts as a query engine that maps relational tables to a virtual graph (Gremlin/Cypher) but highlights why true horizontal partitioning of traversals is impractical. Building a RAG with a knowledge graph + FalkorDB improves task success ~20% over files and another ~20–30% over text search; live LLM-generated ontologies show roughly 50% success without curation.

AI-written summary · about

  1. 0:00Introduction to modern graph databases
  2. 1:13Why graph databases exist
  3. 4:04Storage models in graph databases
  4. 9:00Graph database popularity today
  5. 10:49Neo4j and OLTP graph databases
  6. 11:34Embedded analytical graph databases
  7. 19:49Fraud detection with graph queries
  8. 26:33Graph query engines like PuppyGraph
  9. 39:40Graph databases in AI and context
  10. 48:07Building knowledge graphs for AI
  11. 59:00Q&A and summary
~/search
type to search 6,518 conferences, 143,918 speakers and 192,133 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
← →
on a talk: 5 seconds back / forward
0–9
on a talk: jump to 0%–90% of the recording
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.