deepset-ai/haystack
Open-source AI orchestration framework for building context-engineered, production-ready LLM applications. Design modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation. Built for scalable agents, RAG, multimodal applications, semantic search, and conversational systems.
What it solves
Haystack is an AI orchestration framework designed to build production-ready LLM applications. It solves the complexity of managing retrieval, routing, memory, and generation in a transparent and scalable way, allowing developers to move from experimentation to deployment with confidence.
How it works
It uses a modular architecture where developers design pipelines and agent workflows. These pipelines consist of components for indexing, retrieval, and tool calling, which can be customized or extended with loops, branches, and conditional logic to control how context moves through the system.
Who it’s for
It is built for Python developers building scalable RAG systems, multimodal applications, semantic search, question answering systems, and autonomous agents.
Highlights
- Model- and vendor-agnostic: Supports a wide range of providers including OpenAI, Mistral, Anthropic, Cohere, Hugging Face, Azure OpenAI, and AWS Bedrock.
- Context Engineering: Provides explicit control over how information is retrieved, ranked, filtered, and routed before reaching the model.
- Modular Ecosystem: Offers built-in components for memory and evaluation, with a consistent interface for community-contributed extensions.
- Production-Ready: Supports deployment as REST APIs or MCP servers via Hayhooks.