AI Agent Systems Assistant — Hermes / EA / Agentic Workflows

Starting Rate: $3.50/hour
Full/Part-time to start, with room to grow

I’m looking for someone who is already genuinely interested in and experimenting with **AI agents, Hermes, autonomous workflows, and AI employees.

This is NOT a basic virtual assistant job.

I do not need someone who only knows how to use ChatGPT.

I’m looking for someone who understands—or is already actively learning—the next layer:

Founder / Visionary → Executive Agent (EA) → Specialized AI Employees → Tools → Business Outcomes

I am building a real AI-native operating system for my own company and for a live client business, and I want someone who can grow with me as we build it.

What I Mean by This

I want to be able to give an AI employee a real job, not just a prompt.

For example:

Outcome → Job → Context → Tools → KPI → Authority → Escalation

Then surround that employee with governance:

- what it owns,
- what it is allowed to do,
- what requires approval,
- how its work is verified,
- what gets logged,
- what gets remembered,
- and when it needs to stop and ask a human.

The goal is not to build random agents because agents are cool.

The goal is to give AI real business responsibilities and eventually have an Executive Agent / chief-of-staff layer coordinating those workers.

What We’ll Be Building

Examples may include:

- Setting up and testing Hermes
- Creating specialized AI employees
- Giving agents clear jobs and outcomes
- Building memory and knowledge systems
- Creating reusable skills / SOPs
- Connecting APIs and MCP tools
- Connecting Slack or another messaging interface
- Giving agents limited permissions
- Creating approval and escalation rules
- Testing whether agents actually remember corrections
- Setting up customer/lead workflows
- Working with databases such as Supabase
- Connecting lead sources, email systems, sales systems, and other tools
- Helping AI workers hand complete assignments to human VAs
- Documenting what works and what fails

One real project involves leads entering a business from advertising and building an AI-assisted system that can organize them, segment them, follow up appropriately, surface important responses, and send approval requests to the business owner instead of requiring someone to manually manage every lead.

How I Want You to Think

I am very influenced by this approach:

Do not start with:

“What cool AI tool can we use?”

Start with:

What outcome do we need?

Then determine:

Who owns the job?
What context do they need?
What tools do they need?
What does success look like?
What can they do themselves?
When do they need approval?**

Then let the agent attempt the work.

Test it.

Correct it.

Capture what was learned.

Retest it.

Then improve the system.

I want someone who enjoys figuring things out rather than waiting for a 47-step SOP for something nobody on the team has ever built before.

You Might Be a Fit If You Have Used or Studied Things Like:

- Hermes / Hermes Agent
- Claude Code / Codex / agentic coding tools
- OpenRouter
- MCP
- APIs
- Supabase
- n8n
- GitHub
- Vercel
- Slack bots
- AI memory systems
- Markdown-based agent instructions
- Agent skills
- Multi-agent systems
- Browser/computer-use agents
- AI chief-of-staff / executive-agent architectures

You do NOT have to be an expert in every one of these.

I care much more that you understand the architecture and way of thinking.

Very Important:

Please do not apply if your AI experience is mainly:

- asking ChatGPT questions,
- writing prompts,
- generating captions,
- using Canva AI,
- or watching AI videos on YouTube without actually building anything.

I need someone at least one step ahead of a normal AI power user.

What Success Looks Like

I want someone who can sit beside me virtually while we build this and say things like:

> “Yes, I understand what you are trying to do.”

> “This should probably be a separate employee.”

> “That belongs in memory, not the skill.”

> “This agent should not have permission to do that yet.”

> “We can connect this through an API.”

> “Let’s test this before we make it autonomous.”

> “That result failed. Here is why, and here is what I would change.”

I want a builder and thinking partner—not someone who blindly follows instructions.

Growth Opportunity

I am documenting this entire process while building a company around it.

If you are good, curious, reliable, and genuinely interested in this new way of operating companies, I would love for this to become a long-term role and increase in responsibility and compensation as we grow.

You would be learning and building alongside me very early in this shift.

TO APPLY

Please answer these questions:

1. Have you personally installed or used Hermes Agent? If yes, describe exactly what you built with it.

2. Have you ever created an AI agent that had a real job and could use tools—not just answer questions? Explain what it did.

3. In your own words, what is the difference between:
- a model,
- an agent,
- a skill,
- memory,
- and a tool?

4. If I told you:
“I have Facebook leads entering a spreadsheet and I want the company to intelligently follow up with them without me manually managing every lead,”
what would you think about first?

5. What is MCP, and have you ever used it?

6. Have you worked with any of these?
- Supabase
- APIs
- n8n
- GitHub
- OpenRouter
- Slack bots
- agent memory systems

7. Send me one example of something agentic you have actually built, tested, or attempted—even if it failed.

Start your application with the words: “I know the difference between using AI and deploying AI.”

If that sentence makes immediate sense to you, I want to hear from you.