RAGStack, the Ready-Made RAG Solution, Is Now Integrated with LlamaIndex

Since we introduced RAGStack in November, it’s been exciting and gratifying working with customers who’ve been using this out-of-the-box way to simplify retrieval-augmented generation in AI apps. Today, we're thrilled to announce another significant step in streamlining the RAG implementation process: RAGStack powered by LlamaIndex is now generally available (GA).

What's new?

Implementing RAG on your own can still be complex and overwhelming due to the multitude of choices in orchestration frameworks, vector databases, and large language models (LLMs). On top of new software, RAG itself is changing, with a multitude of new techniques to implement it. 

RAGStack helps surmount these challenges and accelerate GenAI applications on the path to production. With the addition of LlamaIndex to our stack, RAGStack now offers enhanced capabilities in indexing and retrieving data, further streamlining the process of implementing RAG techniques. LlamaIndex brings advanced indexing algorithms and efficient data retrieval mechanisms, making it a perfect complement to RAGStack's existing features and deep integration with DataStax Astra DB. 

Additionally, we're excited to introduce a preview integration of LlamaParse. LlamaParse is LlamaIndex’s new offering that targets enterprise developers building RAG over complex PDFs;  it enables the clean extraction of tables by running recursive retrieval and promises more accurate parsing of the complex docs that enterprises often wrestle with. 

Our commitment to ensure compatibility with LangChain hasn’t changed. RAGStack hasn’t forked LangChain and won’t fork LlamaIndex; each version of RAGStack will be pinned to a particular release of each of the app development frameworks. DataStax will continue to push any fixes or modifications back into both LangChain and LlamaIndex within one quarter. 

Simplifying implementation, amplifying efficiency

Developers can now leverage RAGStack's curated software components alongside LlamaIndex's powerful indexing capabilities, reducing complexity and accelerating the development of production-grade RAG applications. One such example is to use LlamaIndex for data ingestion and LangChain for the query path in a single app to get the best of both worlds. 

A key benefit of integrating LlamaIndex into RAGStack: seamless compatibility with a growing list of embedding models, LLMs, and RAG patterns that enterprise customers use today. You can see the status of these tests live in RAGStack’stest compatibility matrix. Developers can expect improved response times, scalability, and cost savings, all while delivering exceptional AI applications.

RAGStack in action

At DataStax, we're committed to driving AI innovation. By incorporating LlamaIndex into RAGStack, we're staying true to our mission of providing developers with cutting-edge tools and technologies to simplify the complexities of AI application development.

RAGStack benefits from its adoption: it was built to help bring more generative AI applications to production, and as more customers use it, the learnings we gain from them are used to improve RAGStack. 

San Francisco-based Imprompt relies on RAGStack powered by LlamaIndex to power its LLM-driven “chat-to-everything” platform.

 "At Imprompt, we're pioneering digital interaction with our Chat-to-Everything platform, seamlessly integrating with Astra DB from DataStax as our primary vector store,” said Imprompt CEO Jeff Schneider. “With RAGStack powered by LlamaIndex, we enhance enterprise offerings while prioritizing privacy. Our integration with RAGStack enables secure data exchange, empowering innovation for global enterprises."

To learn more about RAGStack and explore the possibilities with LlamaIndex or LlamaParse, visit the documentation. Stay tuned for more updates and enhancements as we strive to make AI application development more accessible and impactful for developers worldwide.

Learn more about LlamaParse during the livestream “Vectorize Your PDFs for GenAI with the LlamaParse API” on March 12. Save your spot!