Summary
This tutorial shows how to set up a personal AI agent that runs around the clock and takes instructions through WhatsApp. The agent is Hermes, an open agent framework that works with models from OpenAI, Anthropic, and Gemini, so users are not tied to a single provider.
The video explains why an always-on agent needs a virtual private server (VPS) rather than a laptop, and walks through a one-click Hermes install on Hostinger, including model selection through the Hostinger AI router. Once the dashboard is open, the presenter covers the four building blocks of the system: memory, which lets the agent retain details such as a user’s niche and writing style across new chats; skills, which are saved instructions that the agent can also write from feedback; channels, including linking WhatsApp by QR code with an allow list; and scheduled jobs, which run tasks at set times such as every morning at 7 a.m.
The demonstration then connects Gmail using an app password so the agent can report which emails need attention. Viewers who are new to AI agents will come away with a clear picture of how an agent differs from a chatbot and how to get a first one running without writing code.
📺 Source: Zinho Automates · Published November 19, 2026
🏷️ Format: Tutorial Demo







