Hezo

Hezo lets you deploy a team of AI agents with defined organizational roles. The homepage describes a hierarchy: a CEO agent, captain agents, and specialist agents, each hired, tuned, and run to deliver specific work. The self-hosted model means your API keys stay yours and the agents run in isolated containers.

The organizational metaphor is more than marketing. Instead of building one general-purpose AI assistant, Hezo creates a structured team where different agents handle different types of work. The CEO agent presumably coordinates and delegates. Captains manage workstreams. Specialists execute specific tasks. This mirrors how human organizations divide work.

The self-hosted positioning is important for security-conscious users. Your agents access your data and systems through your own API keys. No third party sees your prompts, data, or agent configurations. The isolated container architecture means each project runs independently, preventing cross-contamination.

The support for multiple model providers (Claude, GPT, Gemini, local models via Ollama) gives flexibility. The target users are technical teams and businesses that want to automate complex workflows through coordinated AI agents rather than single-purpose automation. The competition includes CrewAI, AutoGen, and LangGraph. Hezo differentiates with the organizational structure metaphor and the self-hosted deployment model.