Live Q&A
summary
Claims that AI is already delivering value while also creating hype and some failed bets. Companies report more than 10 production projects on graphs, driving a push from schema-optional design toward first-class data models and ontologies. Workflow specifics: extract entities from PDFs into a lexical graph then map to domain graphs using LLM hints and embeddings; use language-specific embedding models or fine-tuned BERTs for non-English text. Architecture prescription: model + compute + database, with Neo4j tooling like the knowledge graph builder supporting that stack.
AI-written summary · about
chapters (11)
- 0:00Welcome and introduction
- 1:05AI bubble and industry trends
- 6:01AI in daily engineering work
- 7:47Organizational AI adoption
- 10:13Lessons from Andrew Ng
- 14:09Graph modeling and schema balance
- 20:34Product evolution at Neoforj
- 21:15Multilingual AI challenges
- 24:25Frontier AI research
- 26:15Future of databases and AI
- 30:11Closing remarks
related talks
29mKnowledge graph based chatbots
28mGrounded Intelligence Boosting LLMs with Real-Time Graph Context
32mFireside Q&A
15mTraining and Using Generative AI Models with GraphQL
24mDataEngBytes 2023 - MEL-T2-03 - Joshua Yu
01h 26mThinking in Graphs: The Cognitive Foundations of Agentic AI
41mEnhancing LLMs with Graph Technology
31mWhy AI needs GraphQL