WORKSHOP
DIY RAG: Building a Knowledge Graph-Enhanced Retrieval System with Claude, Neo4j, and Local Vector DBs in Docker (WS04)
27th MAR 2025 | 2:30 PM to 6:00 PM
Venue: Room 1, NIMHANS Convention Centre, Bangalore
FEES: ₹2999
Earlybird price: ₹1999
Objective of the workshop:
This workshop empowers participants to rapidly prototype and deploy RAG applications by providing a hands-on experience in integrating Claude Desktop with local Vector and Graph databases, enabling them to leverage the power of prompt engineering to unlock insights from their data and explore the benefits of advanced techniques like MCP and Agentic RAG

Manjunath Janardhan
Senior Software Engineer and Innovator
- This workshop is perfect for developers, data scientists, and AI enthusiasts who are interested in learning how to build powerful RAG applications. No prior experience with graph databases or vector databases is required. If you're looking to leverage your own data to create intelligent, conversational applications, this workshop is for you!
-
I. Introduction to RAG
Overview of RAG concepts and components -
II. Environment Setup
1. Docker Desktop and Claude Desktop Setup
2. Qdrant and Neo4j Installation with Docker -
III. Integrating MCP
1.Understanding MCP Protocol
2.Setting up the MCP Server for Neo4j with Qdrant
3.Configuring MCP Neo4j and Qdrant -
IV. Building the RAG System
1. Data Preparation and Ingestion
2. Vector Embedding Generation and Storage (Qdrant)
3. Knowledge Graph Creation and Data Loading (Neo4j) -
V. Chatting with Your Data
1. Experimenting with Prompts and Data -
VI. Q&A and Resources
1. Open Discussion
2. Next Steps and Resources
- Hands-on RAG Expertise: Learn by doing and build a functional RAG system using Claude Desktop, Neo4j, and a Vector Database within Docker.
- Unleash Data Potential: Transform your unstructured data into a powerful, searchable knowledge base.
- Master Prompt Engineering: Develop the skills to craft effective prompts for intelligent information retrieval.
- Cutting-Edge Technologies: Gain experience with in-demand technologies like Neo4j, Vector Databases, and Claude for building advanced AI applications.
- Career Ready: Enhance your portfolio and boost your career prospects in the rapidly growing field of AI-powered knowledge management.
- MCP Integration: Gain a basic understanding of how the MCP protocol could be integrated into the Agentic RAG Pipeline.
- Agentic RAG understanding Broad Understanding of framework of RAG.
- A laptop with a stable internet connection
- Basic computer literacy.
- A free Anthropic account for using Claude Desktop.
- *Note that Claude Desktop and Docker Desktop will be a must but I will be providing the setup as part of the workshop.
About Speaker
Manjunath Janardhan is a seasoned Senior Software Engineer with over 20 years of experience transforming healthcare through innovative technology. A patented innovator, Manjunath is a champion of open-source AI and specializes in leveraging these powerful tools to create accessible, scalable, and cost-effective solutions. Currently at GE Healthcare, he’s at the forefront of developing and deploying Gen AI applications with open-source models to enhance developer productivity and accelerate innovation across global teams in the US, China and Japan.
Manjunath is passionate about democratizing AI with open-source projects and has hands-on experience with LLMs, LangChain, Llama-Index, RAG frameworks, and various Agentic Frameworks. He has spearheaded the development of MICT GPT, Code GPT, Service GPT and Unit Test Case Generator to solve real-world problems for developer productivity.
Beyond his technical expertise, Manjunath is a strong communicator and mentor, adept at translating complex concepts into actionable insights. He actively contributes to open-source projects and is committed to driving strategic technology adoption across global healthcare platforms. He offers a unique perspective on the transformative potential of open-source in AI and the importance of community collaboration in building the future of technology