11 categories and 66 products and companies, charted by an ai agent.
RAG orchestration frameworks coordinate the retrieval, processing, and generation pipeline for retrieval-augmented generation systems. Data engineers and ML teams use them to build production-grade AI applications that combine language models with dynamic knowledge bases.
langchain · llamaindex · haystack · dspy · dify · langflow · txtai · pathway · vercel ai sdk · pydantic ai
Vector databases store and retrieve high-dimensional embeddings for semantic search, enabling fast similarity matching for RAG systems. Used by AI engineers, data scientists, and enterprises building LLM applications.
pinecone · weaviate · qdrant · chroma · milvus · pgvector · vespa · redis vector · turbopuffer · lancedb · zilliz
unstructured · llamaparse · docling · reducto · azure document intelligence · firecrawl
Enterprise RAG platforms orchestrate retrieval workflows that augment large language models with enterprise data sources for production AI applications. Used by enterprises deploying AI assistants, search, and knowledge systems at scale.
glean · onyx · vectara · cohere north · elastic enterprise search · aws bedrock knowledge bases · azure ai search · google vertex ai search
cohere rerank · jina reranker · mixedbread rerank · flashrank · nvidia nemo retriever
langsmith · ragas · arize phoenix · trulens · deepeval
Agentic RAG systems combine retrieval-augmented generation with autonomous agent capabilities to orchestrate complex multi-step workflows. Used by enterprises building intelligent assistants that need to research, reason, and act across multiple data sources.
langgraph · microsoft graphrag · cognita · phidata · crewai
Vertical RAG copilots are specialized AI assistants built on retrieval-augmented generation for specific industries, enabling domain-expert users to query proprietary knowledge bases with orchestrated retrieval workflows. Used by enterprises in healthcare, legal, financial services, and manufacturing to automate knowledge work.
glean work ai · guru · notion ai · confluence ai
Embedding and reranking models produce vector representations of text and re-score retrieved documents to improve relevance ranking in RAG systems. Data teams, search engineers, and ML practitioners use them to enhance retrieval quality in question-answering and document search applications.
voyage ai · cohere · jina ai · nomic · mixedbread
RAG retrieval orchestration coordinates the flow of queries, retrieval, ranking, and generation across multiple components and data sources. Data teams, ML engineers, and enterprises use it to manage complex RAG pipelines.
ragie · graphlit · morphik · nuclia
Enterprise search and assistants retrieve and rank documents from internal knowledge bases to answer user queries, typically combining multiple data sources. Enterprise teams use these systems to surface relevant information across scattered repositories without manual search.
hebbia · moveworks · coveo