OpenRAG: An open-source stack for RAG
summary
OpenRAG assembles Docling for PDF‑aware parsing, OpenSearch with the JVector KNN plugin for disk‑based live indexing, and LangFlow for drag‑and‑drop agentic pipelines, and ships as an opinionated open‑source stack (v0.4.0). It runs locally with Ollama models like Granite 4 3B and Queen 3 0.6B embeddings, supports hybrid vector+keyword search, exposes filters and cloud connectors, and uses agentic retrieval to let models decide searches. Visible result: a working end‑to‑end RAG baseline.
AI-written summary · about
chapters (8)
- 0:00The challenges of RAG
- 2:09Introducing Open RAG
- 2:51Document ingestion with Docling
- 5:14Indexing and search architecture
- 7:12Generation and agentic retrieval
- 8:25Open RAG in action
- 12:05Customizing flows in Langflow
- 14:17Conclusion and contributions
related talks
27mElevating AI Applications with OpenSearch's Flow Framework and RAG Tool
25mSet Up an Open Source RAG in 30 Minutes
11mBuilding easily deployable RAG applications with Llamastack
31mBuild an Agentic RAG system using Langchain, Ollama and Milvus
38mBuild AI-Ready Search: Integrating Docling with OpenSearch for Advanced RA...
48mIntegrating Docling With OpenSearch for Advanced RAG and Agentic Applications
50mModular RAG Architectures with Java and Spring AI
22mOrchestrating Multimodal RAG With Agentic Workflows With OpenSearch