Setup & Deployment
How to deploy Hermes Agent on DigitalOcean (Singapore region recommended), connect Telegram or WhatsApp, and configure your initial context prompt.
Hermes Agent for Southeast Asia
I run an autonomous AI agent.
I can set one up for you.
Hermes Agent is an open-source autonomous AI agent you can self-host. I run one in production. It learns my patterns over time and runs for under $20 a month. This is where I publish setup guides, the configs I use, and how to hire me to set yours up.
The basics
What is an autonomous AI agent?
An autonomous AI agent is software that holds context, makes decisions, and acts on its own schedule without step-by-step instructions. Unlike a chatbot you open and close, an autonomous agent remembers your patterns, builds its own reusable skills, and integrates with the tools you already use every day.
The category is growing fast. Gartner predicts that 33% of enterprise software applications will include agentic AI by 2028, up from less than 1% in 2024.
Hermes Agent is one of the few open-source autonomous AI agents you can self-host. It runs on your own server, picks the model that fits each task, and gets sharper over time without retraining anything.
Why this matters
Built for businesses across Southeast Asia.
If you run a business in Southeast Asia, you're probably paying for ChatGPT Plus or Claude Pro and watching your team toggle between five tools they each use a fraction of. The AI lives in someone else's app, doesn't know your business, and starts every conversation from scratch.
The region is investing heavily in AI. Over $2.3 billion went into 680+ AI startups in H1 2025 alone, more than 30% of all private funding in Southeast Asia (Google e-Conomy SEA 2025). The autonomous AI agent category is part of that wave.
An autonomous AI agent is different from a chatbot. It runs on your server, holds your business context permanently, and reaches you in the apps you already use.
What it actually does
Monday morning briefings
Pulls together your weekend emails, the week ahead in your calendar, and what changed in the numbers, delivered before your first coffee.
Customer triage on Telegram and WhatsApp
Handles incoming messages overnight, drafts replies, and flags only the ones that need you. Telegram today. WhatsApp once you give it a number.
Weekly numbers that find you
Pulls revenue, CAC, and pipeline data from your CRM, compares it to the four-week rolling average, and tells you what is off.
A thinking partner that remembers
Joins your strategic thinking and remembers what you decided last quarter, last month, last week. Gets sharper every cycle.
Why self-hosting matters in Southeast Asia
Your data, your server
Hermes Agent runs on infrastructure you control. Important for businesses handling PDPA (Singapore), PDP (Indonesia and Thailand), and DPA (Philippines) regulated data.
No vendor lock-in
Hermes Agent supports any OpenAI-compatible model. Swap providers without rewriting anything. Pick the cheapest model that fits each task.
Fixed monthly cost
Under $20 a month covers the server and the model API for most workloads. No per-seat pricing. No surprise bills as your team grows.
My setup
I run Hermes Agent as my Chief of Staff. Her name is Nora.
"I'm a thinking partner, not a task machine. I know when to just listen."
Nora, from her SOUL.md
Most of what I publish here comes from running her in production across 2Stallions, ChutneyAds, and my angel investments. A full write-up on her setup, the prompt that defines her, and what she has learned is coming soon.
Free Downloads
Configs I Built for Myself
The prompts, setup files, and configs I run on my own Hermes Agent deployment. Drop them into yours and adapt to your needs.
Chief of Staff Prompt Template
Coming soon.
Hermes Agent Setup Checklist (PDF)
Coming soon.
More tools coming as I build them.
Subscribe to get notified.
Guides & Tutorials
Hermes Agent for Southeast Asian Businesses
Operator framing, not developer framing. First-hand experience running Hermes Agent in production at companies of every size.
Use Cases
Real use cases for autonomous AI agents: Chief of Staff, research assistants, customer support on WhatsApp, ops agents, and the workflows people are actually building.
Self-Learning & Memory
How Hermes Agent uses skills and persistent memory to get sharper over time, plus how to audit what your AI agent has learned.
Articles coming soon.
Models, MCPs & Integrations
Choosing the cheapest model that fits each task, extending Hermes Agent with MCP tools, and connecting Telegram, WhatsApp, Slack, and your CRM.
Articles coming soon.
How it compares
Hermes Agent vs ChatGPT Plus vs Claude Pro
How a self-hosted autonomous AI agent compares to the popular SaaS subscriptions most Southeast Asian businesses are already paying for.
| Hermes Agent | ChatGPT Plus | Claude Pro | |
|---|---|---|---|
| Hosting | Your server | OpenAI cloud | Anthropic cloud |
| Monthly cost | Under $20 total | $20 per user | $20 per user |
| Data location | You control | OpenAI controls | Anthropic controls |
| Model flexibility | Any OpenAI-compatible provider | OpenAI only | Anthropic only |
| Self-learning skills | Yes | No | Projects only |
| Telegram / WhatsApp | Yes | No | No |
| Scheduled automations | Yes (built-in cron) | Limited | No |
| Open source | Yes (MIT) | No | No |
Pricing as of May 2026. Verified from OpenAI and Anthropic public pricing pages.
Why two agents?
Hermes Agent works with me. OpenClaw works for the business.
I run an AI cofounder on OpenClaw for business operations. Different jobs, different agents. And two probably isn't the limit either. The question isn't how many agents you run, but how much they free up: work you'd rather not do, work that runs while you sleep, work that would otherwise need another hire.
Frequently asked questions
Autonomous AI agents, answered
What is an autonomous AI agent?
An autonomous AI agent is software that holds context, makes decisions, and runs on its own schedule without step-by-step instructions. Unlike a chatbot that resets every conversation, autonomous agents remember your patterns, build their own reusable skills, and integrate with the tools you already use.
How much does it cost to run a self-hosted AI agent?
Hermes Agent runs on a $6 DigitalOcean droplet plus under $10 a month in model API costs for most workloads. Total: under $20 a month. That's comparable to a single seat of ChatGPT Plus or Claude Pro, except your business gets a persistent agent that serves the whole team rather than one user.
What is the difference between an autonomous AI agent and ChatGPT?
ChatGPT is a chatbot you open and close. An autonomous AI agent like Hermes runs continuously on your server, holds your business context permanently, and acts on its own schedule. ChatGPT forgets your conversation when you close the tab; an autonomous agent remembers and gets sharper over time.
Can I run an AI agent on WhatsApp?
Yes. Hermes Agent supports WhatsApp natively, alongside Telegram, Discord, Slack, Signal, email, and a CLI. You will need a dedicated phone number for the WhatsApp integration. Setup typically takes an afternoon.
Is self-hosting an AI agent secure?
Self-hosting puts you in control of where your data goes. Hermes Agent runs on your own server, talks to whichever model API you choose, and stores its memory on disk you own. For businesses handling PDPA-regulated data in Singapore or PDP-regulated data in Indonesia and Thailand, that control is often the deciding factor.
How long does it take to set up Hermes Agent?
A standalone install takes one afternoon for someone comfortable with SSH. A done-for-you setup with a custom context prompt designed around your business takes one to two weeks, depending on integrations.
Does Hermes Agent work for small businesses, SMEs, and enterprises?
Yes. The same agent platform scales from a solo founder to a 200-person company. The differences are in the context prompt design, the integrations connected, and the workflows the agent handles, not the underlying technology.
Work With Me
Want an autonomous AI agent of your own?
I set up Hermes Agent for businesses in Southeast Asia: startups, SMEs, and enterprises. Server provisioning, Telegram or WhatsApp integration, and a custom context prompt built around your business and how you want your agent to work, whether that's Chief of Staff, research, or operations.
Setup typically takes one to two weeks. You leave with a working agent, not an invoice.
Start My SetupWhat's included
- Server setup . DigitalOcean droplet, Singapore region, configured for production.
- Hermes Agent install . Full agent install, Telegram bot, and swappable model backend.
- Custom context prompt . Designed around your business and the role you want the agent to play.
- Handover & training . You leave knowing how to use, audit, and extend your agent.