feat: add AG2 multi-agent RAG example with Weaviate

README.md

@@ -36,3 +36,4 @@ Weaviate for cool machine-learning related tasks.
 | [Workshop Question Answering](question-answering-application-with-weaviate-workshop) | Python notebook  | Jupyter/Colab notebook to learn how to get started with Question Answering and Weaviate, given at [Knowledge Graph Conference (KGC) 2022](https://www.knowledgegraph.tech/kgc-2022-tutorial-ml-model-with-the-vector-database-weaviate/) |
 | [Attendance system using image2vec-neural and own vectors](attendance-system-example) | Python  |This example uses image2vec-neural and has an option to use own vectors using OpenCV. It is a smart attendance system example|
 | [Monitoring Setup with Prometheus & Grafana](monitoring-prometheus-grafana) | yaml  | This example does not describe any use case, but rather shows a way of how to start, operate and configure Weaviate with Prometheus-Monitoring and a Grafana Instance with some sample dashboards. |
+| [AG2 Multi-Agent RAG with Weaviate](ag2-multiagent-rag) | Python (Jupyter Notebook) | Multi-agent RAG using [AG2](https://ag2.ai/) (formerly AutoGen) with Weaviate as the vector store. AG2 agents collaborate to retrieve and synthesize information via semantic search. |

ag2-multiagent-rag/README.md

@@ -0,0 +1,53 @@
+# AG2 Multi-Agent RAG with Weaviate
+
+This example demonstrates how to use [AG2](https://ag2.ai/) (formerly AutoGen)
+multi-agent conversations with [Weaviate](https://weaviate.io/) for
+Retrieval-Augmented Generation (RAG).
+
+## Overview
+
+Two AG2 agents collaborate to answer questions using documents stored in Weaviate:
+
+- **Research Agent** retrieves relevant documents via Weaviate semantic search
+- **Analyst Agent** synthesizes the information into comprehensive answers
+
+## Prerequisites
+
+- Python >= 3.10
+- OpenAI API key
+
+## Quick Start
+
+```bash
+pip install "ag2[openai]>=0.11.4,<1.0" weaviate-client
+```
+
+Set your API key:
+
+```bash
+export OPENAI_API_KEY="your-api-key"
+```
+
+Run the notebook:
+
+```bash
+jupyter notebook ag2_multiagent_rag_with_weaviate.ipynb
+```
+
+Or open directly in Google Colab:
+
+[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/weaviate/weaviate-examples/blob/main/ag2-multiagent-rag/ag2_multiagent_rag_with_weaviate.ipynb)
+
+## Tech Stack
+
+- [AG2](https://ag2.ai/) — Multi-agent conversation framework (500K+ monthly PyPI downloads)
+- [Weaviate](https://weaviate.io/) — Open-source vector database
+- [Weaviate Embedded](https://weaviate.io/developers/weaviate/connections/connect-embedded) — In-process mode (no Docker needed)
+- [text2vec-openai](https://weaviate.io/developers/weaviate/model-providers/openai/embeddings) — Automatic embedding generation
+
+## How It Works
+
+1. Documents are indexed into Weaviate (auto-embedded via `text2vec-openai`)
+2. AG2 Research Agent uses a registered `search_documents` tool to query Weaviate
+3. AG2 Analyst Agent synthesizes retrieved information into grounded answers
+4. Agents collaborate via AG2 GroupChat with automatic tool execution