The companies building the stack.

Composable pipeline and memory infrastructure for production LLM applications. Enables teams to build, deploy, and scale AI features without managing the underlying orchestration complexity.

Open-source framework for building and deploying production AI agents. Provides the scaffolding — tools, memory, routing — that agent developers need to ship reliably.

Platform for building internal LLM applications and workflows without custom infrastructure. Designed for enterprise teams deploying AI across their operations.

Action execution layer for AI agents — enabling agents to interact with external systems, APIs, and data sources in a structured, reliable way.

Usage-based billing infrastructure for the modern software stack, with native support for AI workload pricing models. The financial plumbing for AI-native businesses.

AI-powered documentation platform that keeps developer docs accurate and current. The surface where developers first encounter a product — built to meet them where they work.

AI system for code review and codebase search — understands the full context of a repository and surfaces relevant patterns, risks, and suggestions at review time.

AI-native knowledge management that organizes information automatically using LLMs, eliminating the manual overhead of traditional note-taking and filing systems. A self-organizing workspace that uses AI to surface relevant context and connections across notes, documents, and conversations. Designed for knowledge workers who need their tools to think alongside them, not just store for them.

Real-time event streaming infrastructure purpose-built for AI agent workflows. Enables agents to react to data changes, trigger downstream actions, and maintain state across long-running tasks.

Enterprise retrieval-augmented generation platform built for production accuracy. Closes the gap between what a foundation model knows and what a business needs it to know.

AI gateway that routes, monitors, and optimizes LLM API calls across providers. Gives engineering teams the reliability, cost control, and observability they need to run AI in production.

Production memory layer for AI agents — enabling agents to maintain context, recall past interactions, and develop persistent knowledge over time. The substrate for agents that get smarter with use.

Secure sandbox infrastructure for AI agents that need to run code — isolating execution environments so agents can write, test, and iterate on code without touching production systems.

Prompt engineering and evaluation platform for production AI applications. Helps teams version, test, and deploy prompts with the same rigor they apply to application code.

Developer platform for building conversational AI agents with access to tools, APIs, and data sources. Abstracts the complexity of agent routing and multi-turn reasoning so teams can focus on product.