Cloud-Native Knowledge Graph Automation to Deliver AI Agents API

Session Track: Applications

Session Time: 04:30 am - 05:00 am November 7

Session description

I'll show how we built an automated, cloud-native continuous integration, delivery, and deployment pipeline for and around Neo4j. The final product is a chatbot that answers complex data-driven questions instead of using human teams. This talk will also focus on how we mixed on-prem and cloud APIs thanks to GitHub Actions and GCP/Google APIGEE. I'll speak about the GenAI stack, which mixes OpenAI and local LLMs with Ollama. Lastly, I'll introduce how we changed our API design based on AI-UX feedback, instead of human feedback, and the paradigm shift it introduces in our daily lives and why it matters.