📄️ Agents
Agents are the core building blocks of Grid applications. They combine language models, tools, and custom logic to create intelligent systems that can understand, reason, and act.
📄️ Tools
Tools are functions that agents can call to perform specific actions, retrieve information, or interact with external systems. They extend an agent's capabilities beyond pure language processing.
📄️ Services Architecture
Grid's services architecture follows a layered design pattern, organizing functionality from atomic building blocks to complex orchestration systems. This architecture ensures modularity, reusability, and clear separation of concerns.
📄️ Conversation Management
Grid provides a sophisticated conversation management system that handles message history, context preservation, and stateful interactions across different conversation patterns.
📄️ Observability
Grid provides comprehensive observability through Langfuse integration, enabling you to monitor, debug, and optimize your AI applications in production.
📄️ Conversation Primitives
Grid provides a hierarchical set of conversation primitives that enable flexible conversation management. Each primitive can be used independently - you don't need to use all layers. All primitives use a closure-based functional pattern - no classes, just functions returning objects with methods.
📄️ Architecture Overview
Grid's conversation primitives follow a layered architecture that provides flexibility and composability at every level.
📄️ Voice Services
Grid's voice services bring natural language interaction to your agents through text-to-speech (TTS) and speech-to-text (STT) capabilities. Voice is implemented as a first-class service following Grid's closure-based architecture pattern.
📄️ Future Roadmap
Learn about upcoming features and the future direction of Grid's conversation primitives.
📄️ Memory System 🚧 Beta
The memory system is currently in beta. APIs may change in future releases.