jaamun

Jaamun implements a schema-based function calling mechanism that allows users to define and invoke functions across multiple AI model providers. This capability utilizes a flexible registry system that maps function signatures to specific APIs, enabling seamless integration with various models like OpenAI and Anthropic. The architecture supports dynamic loading of function definitions, allowing for easy updates and extensions without downtime.

Jaamun provides a contextual state management system that retains user interactions and model responses across different sessions. This is achieved through a lightweight in-memory store that captures the context of interactions, allowing for more coherent and context-aware responses from the AI models. The architecture is designed to minimize latency while maintaining state integrity.

Jaamun enables dynamic orchestration of API calls to various AI models based on user-defined workflows. This capability leverages a rule-based engine that evaluates conditions and triggers API calls accordingly, allowing for complex workflows to be constructed without hardcoding logic. The orchestration engine supports both synchronous and asynchronous operations, providing flexibility in how tasks are executed.