It’s clear that the future of application development will involve AI models and that frameworks must adapt to expose the right abstractions for working with AI. In this episode of Neo4 Live we explore the open source Modus API framework and the building blocks of model-native app development: chaining data and LLMs together, LLM tool use, and knowledge graph RAG. We’ll see how to get started with Modus and Neo4j and add a GraphRAG feature to a Modus + Neo4j model-native app.

Guest: Will Lyon

Modus on GitHub: https://github.com/hypermodeinc/modus
Hypermode: https://hypermode.com/home
Neo4j + Modus blog: https://hypermode.com/blog/neo4j-support-modus
Neo4j + Modus docs: https://docs.hypermode.com/modus/sdk/assemblyscript/neo4j

Slides: https://hyp.foo/neo4j-live-slides
Code: https://hyp.foo/recipes
Full Stack GraphQL Applications: https://www.manning.com/books/fullstack-graphql-applications
Building Knowledge Graphs with LLMs: https://graphacademy.neo4j.com/courses/llm-knowledge-graph-construction/
GraphRAG: https://graphrag.com/

0:00 – Introduction & Welcome to Neo4j Live
6:41 – Understanding AI, Knowledge Graphs & Agentic Workflows
13:08 – Introduction to Modus: A Model-Native API Framework
24:06 – Hands-On: Getting Started with Modus & GraphQL API
33:35 – Building AI-Powered Apps with Function Calling & LLMs
46:02 – Knowledge GraphRAG: Enhancing AI with Neo4j
52:28 – Live Demo: Movie Recommendations with Neo4j & Vector Search
58:48 – Q&A: Use Cases, Next Steps & Community Engagement
1:00:42 – WrapUp

#neo4j #graphdatabase #graphrag #genai #llm #hypermode #modus #appdevelopment