Personal AI Agent Setup for Scaling Products and Non-Profit Operations

Agent-Based Operational Scaling

The author describes their current setup involving six AI agents to manage and scale multiple product ventures and a non-profit organization. This approach aims to reduce the need for extensive human staff and volunteers by automating various operational tasks.

Defined Agent Roles

The six agents are assigned specific profiles: an 'ea-agent' for executive administration, an 'ops-agent' for site performance and issue triage, a 'dev-agent' for core development with GitHub access, a 'gtm-agent' for marketing and social media, a 'research-agent' for web searches and report building, and a 'vps-agent' for infrastructure management with root access. Each agent has distinct responsibilities to cover different aspects of the operations.

Technical Implementation

All agents are currently Hermes agents, defined by SOUL.md, AGENTS.md, skills, tools, MCP servers, and individual Mnemosyne memory banks. A shared Obsidian wiki, synced to the author's machine, serves as a 'business operating manual' for the agents, facilitating knowledge capture and operational consistency.

Early Stages of Experimentation

The author notes that this agent setup is still in its early stages, approximately one month into the implementation. The primary objective is to establish a functional 'staff of agents' to support the growth of their projects.