01What Hermes Agent OS Setup Is and How It Works

The Hermes agent OS setup is a system that makes Hermes AI much more useful. It combines voice commands, agent mode, and memory into one place. You can say a wake word like "Jarvis" or "Hermes" to turn it on.

Then you can ask it questions, give it tasks, or let it listen in the background.

Hermes connects to your Obsidian second brain, which Julian calls the "memory galaxy." Every conversation gets saved there. You can go back later to see what you built. This makes all your AI work easy to find and manage.

The system also works with MCPs, so you can control models and settings from a dashboard.

02Main Features of the Hermes Agent OS

The Hermes agent OS has a mission control area where you see everything you built. You can open past builds and check the log. A paperclip tool lets you create AI agent teams.

There is also an agent group chat where Hermes can talk to other AIs like Claude and Gemini.

  • Voice wake word (Jarvis or Hermes) for hands-free use
  • Agent mode that stays ready in the background
  • Full chat history and a build log for every project
  • Integration with Obsidian memory galaxy for long-term recall
  • MCP management for controlling models and the dashboard
  • Built-in paperclip for AI agent teams and group chat

Julian says this setup makes Hermes feel 100x more powerful. You can type, talk, and build agents all in one system.

03How to Set Up the Hermes Agent OS

To get the full Hermes agent OS setup, Julian offers a zip file inside the AI Profit Boardroom. You also get video tutorials and daily updates. Here are the steps to set it up:

  1. Go to theaiprofitboard.com or click the link in the description.
  2. Join the AI Profit Boardroom community.
  3. Find the agent OS section and download the latest zip file.
  4. Follow the video tutorial to install the prompts and files.
  5. Set up your wake word (Jarvis or Hermes) in the system.
  6. Connect your Obsidian vault if you want the memory galaxy feature.
  7. Test voice commands like "Jarvis, show me my numbers."

If you prefer a free option, you can use Claude Code with Qwen 2.5 Coder and Ollama on your local machine. That runs on a free local model.

04How to Use the Hermes Agent OS: Examples Shown

Julian shows several ways to use the Hermes agent OS. First, he says "Jarvis, build me a galaxy I can swirl with my mouse." The system builds it and responds. Then he asks "Jarvis, what have I been working on this week?" Jarvis checks the thumbnail log and says Julian worked on thumbnails.

You can also ask for jokes. Julian says "Jarvis, tell me a joke that's about AI." The neural network joke is a fun example. For business, he says "Jarvis, show me my numbers" to pull up data.

He even opens Julian Goldie's site by voice. To view all builds, he says "Show me everything I've built" and sees a list of past projects.

Another feature is the agent group chat. Hermes can talk to Claude and Gemini in one place. Julian also combines Hermes with 11 Labs for voice output, with Claude helping in the background.

05Results and Comparisons from the Episode

Julian compares the Hermes agent OS to a free setup using Claude Code. The free version uses Qwen 2.5 Coder with Ollama on a local model. Julian says local models are getting better and better.

He calls the Hermes setup his favorite part.

He mentions that the agent OS makes it easier to manage everything in one place. The memory galaxy logs all conversations, so nothing is lost. Julian thinks this might be the future of AI.

For getting stuff done, he suggests the Hermes setup or Paperclip can create more a lot faster.

The free Claude Code alternative has its own agent factory. You can speak or type to it, and it builds things live. Julian shows some examples of what he built with it.

But he clearly prefers the Hermes agent OS for its completeness.

06Tips, Limits and Mistakes to Avoid

One tip Julian shares is to use the wake word properly. Say "Jarvis" or "Hermes" to turn the system on. You can switch it off when not needed.

This saves battery and avoids mistakes. He also recommends keeping your Obsidian memory linked so you never lose past work.

  • Always test voice commands first to make sure the wake word works.
  • Don't forget to check the build log – it shows everything you created before.
  • If you use the free Claude Code version, make sure Ollama is installed correctly.
  • Limits: local models may be slower than cloud models.
  • Mistake to avoid: expecting the system to work instantly without setup.

Julian says the community inside AI Profit Boardroom helps you avoid these mistakes. You can ask questions and get help 24/7. The tutorials are updated daily too.

07Who the Hermes Agent OS Is For

The Hermes agent OS setup is for anyone who wants to build with AI faster. Julian created it for people in the AI Profit Boardroom community. That group includes learners, scale-ups, and people growing with AI automation.

If you enjoy party tricks and futuristic demos, this setup is fun to explore.

It matters because it shows how voice, memory, and agents can work together. Julian believes this might be the future of how we interact with AI. The setup also includes weekly coaching calls, a map to connect locally, and real-time support.

Whether you are a beginner or advanced user, you can learn and build with it.