Building Multi-Agent AI Systems with Whisperer Framework in Elixir

Whisperer is a framework designed for building and orchestrating multi-agent AI systems, developed by MonitorLizard, an observability platform that utilizes Opentelemetry and Agentic AI. The need for a straightforward method to construct and manage observability-based agents, which assist teams in resolving production issues, led to the framework's creation. Whisperer is built around a set of behaviors for defining agents and includes a fault-tolerant orchestrator for executing these agents using large language models (LLMs). Unlike a traditional library, Whisperer allows users to choose their implementations and sequencing strategies, simplifying the orchestration process while remaining adaptable to changes in LLMs and Agentic systems. This makes Whisperer a reliable framework for leveraging innovations in the agent-based AI field without compromising core stability.