DataWeave Introduces Open-Source Framework for Building Retrieval-Augmented Generation Pipelines

SAN FRANCISCO, California, April 8, 2026 /Pitchwire/ /Pitchwire/ — DataWeave, a developer tools company, today announced the open-source release of RAGsmith, a modular framework for building production-grade Retrieval-Augmented Generation (RAG) pipelines.

RAGsmith provides a standardized architecture for connecting large language models to enterprise knowledge bases, with built-in support for chunking strategies, embedding models, vector stores, reranking, and evaluation. The framework supports 15 vector databases, 8 embedding providers, and any LLM accessible via OpenAI-compatible APIs.

"Every company building RAG is reinventing the same wheel — chunking, embedding, retrieval, reranking, evaluation," said Anika Patel, CTO of DataWeave. "RAGsmith gives teams a production-ready scaffold so they can focus on what actually differentiates their application: the domain-specific tuning and the user experience."

The framework has been in private beta with 40 enterprise development teams since December 2025. Contributors include engineers from Stripe, Databricks, and Anthropic.

RAGsmith is released under the Apache 2.0 license and is available at github.com/dataweave/ragsmith. DataWeave offers a managed cloud version with monitoring, A/B testing, and enterprise support starting at $500/month.

The company has raised $18 million in total funding, including a $14 million Series A led by Andreessen Horowitz in November 2025.