Supercharging Gen-AI with Graphs: From Accurate LLMs to Drug Discovery

Join us in Tokyo for an evening where cutting-edge Generative AI meets the power of connected data. This meetup brings together experts exploring how graphs are transforming AI applications—making them more factual, scalable, and impactful in the real world.

Agenda:
* 5:40pm~6:00pm: Doors open, Checkin and networking
* 6:00pm~6:10pm: Implem
* 6:10pm~6:40pm: First Talk
* 6:40pm~7:10pm: Second Talk
* 7:10pm~7:40pm: Third Talk
* 7:40pm~8:00pm: Open discussion & Mixer, Doors Close

✨ What to expect:

• Learn how GraphRAG powers up LLMs to be more factual and reliable.
• Get introduced to Neo4j MCP Servers and how they enable seamless, AI-ready infrastructure.
• Discover how Graphs and Gemini EQA are accelerating the drug review process in Taiwan, showcasing the potential of AI in healthcare.

Topics/Speakers:
🎤 Sixing Huang, Bioinformatician and Evangelist, Gemini Data Inc.
Talk Topic: Accelerate the Drug Review Process in Taiwan with Graphs and Gemini EQA
Talk Description: Since May 2024, our team has been collaborating with a Taiwanese drug review agency to streamline their new drug application process. The agency faced a significant challenge: efficiently sifting through a vast repository of both internal historical cases and extensive external documentation from international counterparts like the FDA, EMA, and PDMA.
To address this, we developed and implemented a Neo4j knowledge graph. This graph seamlessly integrates both the agency's private case data and the relevant public documents, creating a unified and interconnected knowledge base. To make this information easily accessible, we also provided them with Gemini EQA, an LLM-powered tool designed for intuitive querying of the graph.
Our solution has dramatically improved the agency's workflow. They can now:

• Efficiently find remotely related old cases, accelerating their review process.
• Conduct rapid statistical analyses on drug data.
• Visualize the entire life-cycles of different drugs for better insights.

By connecting these disparate data sources, our work has empowered the agency to make more informed decisions and significantly enhance the speed and accuracy of their drug reviews. In this talk, I would like to give the audience a brief walk-through of the project.
Speaker Bio: To be added shortly

🎤 Siddhant Agarwal, Developer Relations Lead APAC, Neo4j
Talk Topic: Power up LLMs to be Factual Ninjas using GraphRAG
Talk Description: In the age of big data, Large Language Models (LLMs) are powerful yet limited by challenges like factual inaccuracy, poor reasoning, and difficulty with complex information. This talk introduces GraphRAG, a game-changing approach that integrates knowledge graphs with LLMs to overcome these limitations. Unlike traditional LLMs or vector-only RAG systems, GraphRAG leverages structured data to enhance accuracy, enable deeper reasoning, and unlock advanced contextual understanding. Through real-world examples, we’ll explore how GraphRAG bridges the gap between raw data and meaningful insights, showcasing its potential to revolutionize intelligent applications like question answering and chatbots.
Speaker Bio: Siddhant Agarwal is a seasoned DevRel professional with over a decade of experience building and scaling developer ecosystems globally. Currently leading Developer Relations across APAC at Neo4j and a Google Developer Expert in Gen-AI, Sid is known for his unique ability to tell powerful stories through technology—translating complex ideas into compelling narratives that resonate with developers. With his signature "Local to Global" approach, he turns grassroots innovation into global impact. Previously at Google managing flagship developer programs, he has shared his technical expertise at diverse forums worldwide, fueling inspiration and innovation. Know more at meetsid.dev.

🎤 Jason Koo, Developer Advocate, Neo4j
Talk Topic: Introduction to Neo4j MCP Servers
Talk Description: Large Language Models (LLMs) are powerful, but they often struggle to connect seamlessly with structured databases, limiting their ability to retrieve and reason over complex, real-world information. This talk introduces Neo4j MCP Servers, a breakthrough set of tools that enable LLMs to directly interface with Neo4j graph databases. Unlike conventional database connectors or ad-hoc API integrations, MCP Servers provide a standardized protocol for querying, modeling, and managing graph data in ways that are optimized for AI agents. Through practical demonstrations, we’ll introduce how to setup and get started with these MCP Servers.
Speaker Bio: Jason Koo is a Developer Advocate at Neo4j and longtime community builder passionate about helping developers succeed with emerging technologies. A former mobile developer turned Pythonista, he creates demos, talks, and hands-on workshops that bring graph databases to life for developers worldwide. Jason has previously championed developer communities at Sendbird, alwaysAI, and Tealium, producing instructional content, leading meetups, and speaking at conferences.

Whether you’re a developer, data scientist, or simply curious about the intersection of graphs and Gen-AI, this event will give you practical insights and real-world use cases to supercharge your AI journey.