Two core concepts power everything:
1
Agent Management Studio
Your entire product team uses this to build, manage, and optimize agents. It handles data, configuration, tools, and MCP servers. Most importantly, it provides real-time data, allowing you to understand what’s happening in your system. Everything we believe works better as configuration over code lives here. We’re deliberate about what belongs here- where configuration should stay in code, we keep it in code. What lives in the AMS empowers non-developers who are just as crucial to the project and take ongoing responsibility for managing and optimizing your AI systems.
- Business intelligence (BI dashboards, analytics, and reporting)
 - Agent configuration (data sources, tools, and workflows)
 - Agent management (create, update, and delete agents)
 - Agent optimization (real-time performance data and insights)
 - Agent deployment (deploy agents to production environments)
 - Tool integration (connect to external tools and services)
 - Workflow management (define and manage agent workflows)
 - Knowledge management (manage knowledge bases and data sources)
 - Bias identification (identify and mitigate biases in AI systems)
 - Security and compliance (ensure data security and compliance with regulations)
 - Agent policy management (define and enforce policies for agent behavior)
 - Agent outcome tracking (track agent performance and outcomes)
 
2
Developer SDK/APIs
Everything your development team needs to create outstanding user experiences. You’re not locked into our components. We provide a world-class multi-media, multi-modal Agent Communication experience (aka chatbot), but you don’t have to use it. We offer headless APIs and structured data streams to sync agent state between front-end and back-end. You can integrate with any orchestration system or UI component library. We extend your existing stack rather than replace it.
- Agent Communication (chatbot, voice, and multi-modal interfaces)
 - Agent UI configuration (customizable UI components for agents)
 - Agent SDK (for building custom agent experiences, colors and fonts)
 - Headless APIs (for integrating agents into existing applications)
 - Structured data streams (for syncing agent state)
 - Orchestration system integration (for connecting agents to external systems)
 - Front-end and back-end integration (for seamless agent experiences)
 - Real-time data streaming (for live updates and interactions)
 - Customizable agent interfaces (for tailored user experiences)
 - Agent session management (for handling user sessions and interactions)
 - Agent state management (for maintaining agent context and state)
 - Agent interaction history (for tracking user-agent interactions)
 - Knowledge base integration (for connecting agents to knowledge sources)
 - Agent multi-tenantcy (for managing multiple agents across different tennants)