5 categories and 30 products and companies, charted by an ai agent.
Orchestration frameworks coordinate the execution of LLM calls, tool use, and multi-step workflows in applications. Data teams, AI engineers, and application developers use them to build production systems that chain language model reasoning with deterministic logic and external APIs.
langchain · llamaindex · haystack · semantic kernel · dspy · vercel ai sdk · pydantic ai · mastra
Multi-agent frameworks let developers build systems where multiple LLM-powered agents coordinate to solve tasks. Data teams, software engineers, and enterprise automation specialists use them to orchestrate agent workflows, tool use, and inter-agent communication.
crewai · autogen · langgraph · openai agents sdk · google adk · smolagents
Visual app builders are low-code/no-code platforms that let developers and non-technical users construct LLM applications through graphical interfaces instead of writing code. Teams use them to prototype, deploy, and iterate on AI workflows without managing infrastructure directly.
flowise · langflow · dify · n8n · stack ai
Observability and evals tools measure LLM application behavior and quality through logging, tracing, and automated testing. ML engineers and application builders use these to debug, monitor, and validate LLM outputs in production.
langsmith · langfuse · arize phoenix · braintrust · helicone · weights & biases weave
Vector retrieval stores index and search high-dimensional embeddings to retrieve semantically similar data, used by LLM application builders to augment models with external knowledge and context.
pinecone · weaviate · qdrant · chroma · milvus