Your Hermes agent shouldn't disappear when you close your laptop.
To answer a Slack message tomorrow morning, Hermes needs more than a local install: a place to run, a secure web address, and settings that survive updates. This guide shows how to keep Hermes running 24/7 in your own cloud without spending time on each part by hand.
Keep Hermes running when your laptop is off
Hermes is an autonomous AI agent that can work on code, browse the web, remember context across tasks, automate repetitive work, and respond through Slack, Discord, Telegram, and WhatsApp. A local Hermes process only stays available while your computer is awake and connected.
Running Hermes in the cloud - on a computer you rent from a provider - keeps it available without depending on your laptop, Mac mini, home server, or home internet connection. It can use a web address your team knows and keep its settings when you update it.
That takes time: someone has to create that cloud computer, configure the Linux server, connect the web address, protect the dashboard, save its data, and show you what happened if a deployment fails. The Hermes template in Devopness sets up those common parts for you.
Ask your AI assistant to deploy Hermes
If you use Cursor, Claude Code, Codex, or another AI coding tool, you can deploy Hermes from the same conversation. Devopness connects to these tools through MCP, which lets an AI assistant perform approved actions in other products.
- Create a free Devopness account
- Connect the Devopness MCP server to your AI assistant
- Tell your AI assistant:
Use Devopness to deploy the Hermes AI Agent template to my cloud account.
Guide me through the missing details, one step at a time.
Deploy the app and verify that Hermes is running.
Your assistant will ask which cloud account to use and guide you through the remaining choices. If Devopness needs permission to access GitHub or your cloud provider, the assistant will tell you how to connect the account rather than guessing.
Devopness keeps the setup visible. The assistant performs the steps, and you can review what it created in the web app without sharing your cloud keys in the chat.
Use the browser with the same template
You can also deploy in the Devopness web app:
- Open Applications and select Add application
- Choose Self-host an open source app
- Select Hermes AI Agent
- Continue to Configure
The next screen asks you to select your GitHub connection, name the application, enter the web address for the dashboard, and choose a password and a separate random secret to protect its login.
When you confirm, Devopness creates the Hermes application and opens its deployment screen. This is where you choose the server where Hermes will run.
On the deployment screen:
- Keep the default
mainbranch, or choose another version of Hermes - Open Servers and select one or more servers where Hermes should run
- If you don't have a server yet, select Add server and follow the guided setup
- Review your choices and select Deploy
If you're adding a server, choose one with at least 4GB of memory. Browser tasks can use a significant amount of memory.
After you select Deploy, Devopness opens the action page with deployment logs. You can follow every step from a desktop or mobile browser, wherever you are, without connecting through SSH or using Linux commands.
If a step fails, ask your AI assistant to read the logs, explain what went wrong, and guide you through the fix before deploying again.
What the template sets up for you
The template handles these parts:
- connects the application to the official Hermes source code
- tells the server how to start Hermes
- gives Hermes a safe place to keep its settings and data
- protects the dashboard with the login you chose
- connects your web address to the Hermes dashboard
Hermes runs in your cloud account, so you can use any eligible startup or trial credits available from your provider. You choose the servers, web address, connected accounts, and AI model provider. Everything the template creates remains visible and editable in Devopness.
Once Hermes is configured, you can reuse that setup to deploy it to more servers across AWS, Azure, GCP, DigitalOcean, and Hetzner. If you change cloud providers later, you can reuse the setup instead of starting again.
Connect the dashboard and give Hermes a model
At the company where you bought your domain, point the domain to your server's IP address. This is usually called adding a DNS record. Devopness has already prepared Hermes to respond at that address.
Before sharing the dashboard, add an SSL certificate in Devopness. The certificate enables HTTPS and displays the lock icon in the browser. Your password protects the login. If many people will use the dashboard, review the stronger sign-in options in the Hermes dashboard documentation.
Hermes also needs an AI model. Open the application's settings in Devopness and add the API key from your chosen provider: OpenAI, Anthropic, OpenRouter, or Groq. An API key is the private code that allows Hermes to use your provider account. Keep it secret.
The API key and your Hermes settings remain available when you update Hermes. You can then connect Hermes to Slack, Discord, Telegram, WhatsApp, or another supported messaging tool.
Close the laptop. Hermes should keep answering
Before relying on the agent, check four things:
- the Hermes login page loads over HTTPS
- the dashboard accepts the password you configured
- Hermes answers using the AI model you selected
- its settings remain after you deploy again
Now you can use Hermes from the dashboard. That is the fastest way to confirm everything is working and Hermes is answering.
When you are ready to use it through chat, connect Slack, WhatsApp, Telegram, or Discord and message Hermes there too.
Create a free Devopness account and choose the path you prefer: send the prompt above to your AI assistant, or select Self-host an open source app in the web app.
Less setup. More time for the work that matters. Save time. Keep control.
Happy deploying!