Founding Hermes AI Systems Engineer — $2K–$3K/Month (Hiring in 48 hours)

Please login or register as jobseeker to apply for this job.

TYPE OF WORK

Full Time

SALARY

$2,000 - $3,000 per month

HOURS PER WEEK

50

DATE UPDATED

Aug 5, 2026

JOB OVERVIEW

FOUNDING HERMES AI SYSTEMS ENGINEER & DEPLOYMENT LEAD

This is not a normal automation job.

We are building an AI operating system that will first run our own companies and then be installed inside businesses across the United States.

Increase ROAS was one of the fastest-growing agencies in our market in 2025. We have already built successful marketing systems for service businesses, generated substantial client results and proven that we can sell.

Our next stage is bigger:

We want to become the leading company installing practical AI operating systems inside American businesses.

We are hiring the founding technical builder who will help make that happen.

THE MISSION

Your mission is to turn Hermes into a dependable AI operating system that can:

- Monitor a company
- Understand what is happening
- Organize information
- Identify problems
- Prepare decisions
- Assign or recommend work
- Monitor employees and outcomes
- Manage projects
- Run scheduled checks
- Draft communications
- Create and improve SOPs
- Connect business systems
- Escalate exceptions
- Help the business operate without depending on the owner for every decision

We will first implement this system inside:

- Increase ROAS
- Hot Tub Launch
- Start Scale Automate / Aila

After it works internally, we will package and install it inside businesses throughout the United States.

This is not a one-time internal automation project.

You will help create the deployment system, skill library, documentation, security model, monitoring, update process and client-installation framework that can eventually support hundreds of businesses.

CURRENT HERMES ENVIRONMENT

Hermes currently runs on a Mac Mini and is accessed through Telegram.

The environment includes or is intended to include:

- Hermes Agent
- Telegram
- Obsidian memory and RAG
- Terminal access
- Browser automation
- ClickUp
- GoHighLevel
- Meta Ads
- Gmail
- Google Calendar
- Google Drive
- GitHub
- Slack
- Cloudflare
- Supabase
- 1Password
- APIs
- MCP servers
- Cron jobs
- Skills
- Sub-agents
- Approval workflows
- Backups and monitoring

Some components are connected.

Some are partially built.

Some need to be tested, fixed, documented or redesigned.

You must be able to enter an imperfect environment, determine the current truth and finish working systems quickly.

NON-NEGOTIABLE HERMES REQUIREMENT

You must already understand Hermes Agent.

Strong applicants will have personally installed, configured or operated Hermes and will understand concepts such as:

- Hermes installation and upgrades
- Profiles and configuration
- Telegram gateway
- Memory
- Skills
- Cron jobs
- MCP integrations
- Tool permissions
- Terminal backends
- Sub-agents
- Context files
- Logs
- Backups
- Approval controls
- Model routing

It is not enough to have briefly heard of Hermes.

Prior production or personal Hermes setup is strongly preferred.

We will also consider someone who has studied Hermes deeply and built equivalent agent systems but was prevented from completing a full Hermes deployment because of provider, payment or geographic restrictions.

That person must:

- Explain the exact restriction
- Demonstrate equivalent technical work
- Show that they understand the Hermes architecture
- Pass a paid Hermes implementation test

We can provide approved remote access to a US-based Mac Mini and the systems required for the role.

WHAT YOU WILL BUILD

1. INTERNAL AI OPERATING SYSTEM

Turn Hermes into the intelligence and coordination layer for our own businesses.

Hermes should eventually help answer:

- What should the owners focus on today?
- Which clients need attention?
- Which employees are off-track?
- Which campaigns have problems?
- Which projects are blocked?
- Which payments failed?
- Which contracts are missing?
- What is waiting for approval?
- Which system contains conflicting information?
- What should happen next?

2. SCHEDULED BUSINESS WORKFLOWS

Build reliable scheduled workflows such as:

- Daily owner brief
- Employee activity and output report
- Client-risk report
- Campaign anomaly report
- GoHighLevel lead-routing check
- Weekly financial review
- Active-project checkpoint report
- Contract verification
- Billing exception report
- Source-of-truth consistency check
- Strategy-session capture
- Backups and recovery checks

3. HERMES SKILL LIBRARY

Build reusable skills for:

- Operations monitoring
- Project management
- Employee accountability
- Client-risk detection
- Meta Ads analysis
- GoHighLevel analysis
- Finance preparation
- Contract verification
- Task creation
- SOP drafting
- Website QA
- Technical troubleshooting
- Source-of-truth maintenance
- Client onboarding
- Reporting

4. CLIENT DEPLOYMENT SYSTEM

Create a repeatable process for installing a secure Hermes operating system inside other businesses.

The deployment system must cover:

- Discovery
- Business process mapping
- Instance setup
- Memory initialization
- Integrations
- Skills
- Permissions
- Scheduled workflows
- Data isolation
- Security
- Human approval rules
- Testing
- Training
- Monitoring
- Updates
- Backups
- Support
- Offboarding

The goal is not to manually rebuild every client.

The goal is to create a deployment factory.

5. INTERNAL OPERATIONS

Help simplify and connect:

- ClickUp
- GoHighLevel
- Meta Ads
- Email
- Calendar
- Google Drive
- GitHub
- Cloudflare
- Supabase
- Obsidian
- Telegram
- Payment and finance reporting

You will help establish:

- One source of truth for every business object
- Clear system boundaries
- Reliable task creation
- Clear owners
- Due dates
- Definitions of done
- Evidence
- Escalation rules
- Project checkpoints

6. TECHNICAL FULFILLMENT

Help build and maintain systems involving:

- Automated client onboarding
- Intake forms
- Supabase data
- GoHighLevel sub-accounts
- Website generation
- Cloudflare deployments
- Environment variables and secrets
- Meta Pixel and Conversions API
- Lead routing
- Workflow testing
- Client dashboards
- AI voice and messaging systems

FIRST 30 DAYS

Your first 30 days will focus on proving that you can finish systems safely and quickly.

Required outcomes include:

1. HERMES ENVIRONMENT AUDIT

Produce a verified matrix showing every system as:

- Connected
- Authenticated
- Read tested
- Write tested
- Scheduled workflow tested
- Production safe
- Broken
- Not configured

2. DAILY OWNER BRIEF

Build and prove an automated 11: ---------- AM Eastern Time Telegram report containing:

- Top owner decisions
- Clients requiring attention
- Employee priorities
- Overdue work
- Blocked work
- Campaign anomalies
- Unanswered communication
- Active-project checkpoints
- Work waiting for approval
- Source-of-truth conflicts

3. PROJECT CONTROL SYSTEM

Create a dependable project-control process showing:

- Project
- Business
- Desired outcome
- Human owner
- Current state
- Last verified checkpoint
- Next action
- Due date
- Blocker
- Dependency
- Evidence
- Definition of done
- Stop condition

4. CLICKUP EXECUTION SYSTEM

Help create a minimal employee work queue where each person can immediately see:

- What to do
- Priority
- Owner
- Due date
- Definition of done
- Required evidence
- Blocker
- Escalation rule

Do not delete or reorganize historical data without approval.

5. FIRST PROVEN HERMES WORKFLOWS

Deploy and test the first high-value workflows:

- Daily owner brief
- Employee activity report
- Client-risk monitor
- Active-project checkpoint
- Source-of-truth conflict detection
- Strategy-session capture

6. DOCUMENTATION AND RECOVERY

Create:

- Architecture map
- Integration registry
- Skill registry
- Cron registry
- Access matrix
- Change log
- Backup process
- Restore process
- Error-handling process
- Rollback procedures

DAYS 31–60

- Complete internal Hermes operating workflows
- Finish the source-of-truth architecture
- Expand employee and client monitoring
- Help complete client onboarding automation
- Help complete the website-generation pipeline
- Build technical QA and escalation systems
- Convert repeated work into skills, scripts or documented processes

DAYS 61–90

- Build the reusable Hermes client deployment package
- Create standard configurations and templates
- Establish client data-isolation rules
- Create deployment QA
- Create upgrade and maintenance procedures
- Help install the system inside the first external businesses
- Measure installation time, support burden and business impact
- Improve the deployment process based on real failures

REQUIRED TECHNICAL ABILITY

You must be strong in most of the following:

- Hermes Agent
- Python
- JavaScript or TypeScript
- REST APIs
- Webhooks
- OAuth
- JSON
- Databases
- Git and GitHub
- Linux or macOS terminal
- SSH
- Cron and scheduled jobs
- MCP
- Telegram bots
- Cloudflare Pages and Workers
- Wrangler
- Supabase or Postgres
- Environment variables
- Secret management
- Browser automation
- Logging
- Monitoring
- Error handling
- Testing
- Rollback design

Strong preference for practical experience with:

- GoHighLevel
- ClickUp API
- Meta Marketing API
- 1Password
- Composio
- n8n
- Make
- Cursor
- Codex
- Claude Code
- RAG systems
- Obsidian
- Multi-agent systems

ORGANIZATION IS NON-NEGOTIABLE

You must be extremely organized.

Every project you touch must have:

- A clear outcome
- One human owner
- Current status
- Next action
- Due date
- Blocker
- Definition of done
- Evidence
- Documentation
- Canonical files
- Rollback where required

You must:

- Keep tasks current
- Update project checkpoints
- Document decisions
- Use version control
- Attach evidence
- Name files clearly
- Record what changed
- Surface risks early
- Close completed work
- Preserve source material
- Avoid duplicate systems

We do not accept:

- “I am still working on it” with no evidence
- Missing status updates
- Work hidden in private chats
- Unexplained delays
- Undocumented production changes
- Random files with unclear names
- Confident guesses
- Waiting until a deadline passes to report a blocker

WORK SPEED AND EXPECTATIONS

This is a fast-moving founder-led environment.

You must be comfortable with:

- Fast deadlines
- Short build cycles
- Weekly releases
- Multiple systems
- Direct feedback
- Rapid iteration
- Urgent client situations
- Occasional extended build sprints
- Changing priorities when real business risk appears

You should be the type of person who:

- Wants to build something significant
- Takes ownership
- Moves without constant supervision
- Works through difficult technical problems
- Investigates before asking basic questions
- Communicates early when blocked
- Does not disappear
- Does not make excuses
- Does not wait for perfect instructions
- Can turn an unclear problem into a structured plan
- Wants responsibility and long-term growth

This is not a low-pressure maintenance role.

This is also not permission to make reckless changes.

Speed must include:

- Testing
- Security
- Documentation
- Evidence
- Rollback
- Correct client identification
- Approval for production-risk actions

EXTENDED HOURS

The standard schedule is 40 hours per week.

During important client launches or critical infrastructure sprints, we may agree on temporary extended hours.

Extended hours must be:

- Discussed in advance
- Focused on a defined outcome
- Time-limited
- Compensated
- Followed by a return to the normal schedule

We are looking for commitment and urgency, not disorganization or permanent burnout.

THIS ROLE IS NOT FOR YOU IF

- You are a general virtual assistant
- You are only a prompt writer
- You only know basic Zapier automations
- You have never studied Hermes
- You cannot use the terminal
- You blindly run AI-generated commands
- You require step-by-step instructions for everything
- You avoid documentation
- You regularly miss deadlines
- You hide blockers
- You are disorganized
- You make production changes without approval
- You do not understand security
- You do not test your work
- You cannot explain your architecture
- You want a comfortable role with slow deadlines
- You are looking for several simultaneous jobs
- You are not interested in building something with national scale

WHY THIS ROLE MATTERS

You will not simply be maintaining automations.

You will help build the infrastructure behind a company whose goal is to bring AI operating systems into businesses across the United States.

The right person can grow into:

- Head of AI Systems
- Technical deployment leader
- Team manager
- Product architect
- Client implementation leader

You will have direct access to the founders, real businesses, real systems, real client problems and the opportunity to help create a new category.

COMPENSATION

- $2,000–$3,000 USD per month based on demonstrated ability
- 30-day paid trial
- Milestone bonuses for completed production systems
- Performance reviews
- Long-term growth
- Potential leadership opportunity
- Software and approved infrastructure provided

HOW TO APPLY

Start your application with the exact phrase:

HERMES BUILDER

Applications that do not begin with that phrase will not be reviewed.

Include:

1. HERMES EXPERIENCE

Describe your direct Hermes experience.

State:

- Whether you installed it
- Where you installed it
- Which version
- Which messaging gateway you used
- Which models you connected
- Which MCPs you connected
- Which skills you created
- Which cron jobs you created
- Which memory system you used
- Whether you used sub-agents
- What failed
- How you fixed it

Provide screenshots, a Loom video, repository or other evidence.

2. EQUIVALENT AGENT EXPERIENCE

Show one real AI-agent or automation system you personally built.

Explain:

- Business problem
- Architecture
- Tools
- APIs
- Memory
- Permissions
- Error handling
- Testing
- Monitoring
- Security
- Result

3. HERMES ARCHITECTURE QUESTION

Explain how you would design Hermes to:

- Read ClickUp, GHL, Meta, Gmail and Calendar
- Produce a daily owner brief
- Store durable memory
- Create tasks only after approval
- Log every action
- Handle conflicting data
- Stop safely when a system fails

4. CLOUDFLARE SCENARIO

Cursor tells a nontechnical owner to open Terminal and add a backend secret to Cloudflare.

Explain:

- How you classify the risk
- Whether Cursor is the correct operator
- Whether MCP is needed
- How the secret should be handled
- How the change is approved
- How it is verified
- How it is rolled back
- What gets documented

5. SEVEN-DAY PLAN

Provide a seven-day plan for auditing an existing Hermes installation and deploying the first reliable scheduled business report.

6. PERSONAL INFORMATION

State:

- Country
- Time zone
- Eastern Time overlap
- Desired monthly compensation
- Current employment
- Availability
- Earliest start date
- Internet backup
- Power backup
- Whether you can commit exclusively to this full-time role

PAID TECHNICAL TEST

Qualified candidates will complete a paid test in a staging environment.

The test may include:

- Inspecting an existing Hermes installation
- Connecting a mock or test MCP
- Creating one Hermes skill
- Creating one scheduled job
- Sending a Telegram report
- Reading data from an API
- Handling a simulated failure
- Documenting the architecture
- Creating a rollback plan
- Producing evidence

The test will be scored on:

- Hermes knowledge
- Diagnostic ability
- Technical execution
- Speed
- Organization
- Security
- Testing
- Documentation
- Communication
- Escalation judgment

This is a founding-builder position for someone who wants to help build the AI operating system used by businesses across the United States.

VIEW OTHER JOB POSTS FROM:
SHARE THIS POST
facebook linkedin
  BENCHMARKS  
Loading Time: Base Classes  0.0009
Controller Execution Time ( Jobseekers / Job )  0.0241
Total Execution Time  0.0259
  GET DATA  
No GET data exists
  MEMORY USAGE  
1,588,664 bytes
  POST DATA  
No POST data exists
  URI STRING  
jobseekers/job/Founding-Hermes-AI-Systems-Engineer-2K3KMonth-Hiring-in-48-hours-1703419
  CLASS/METHOD  
jobseekers/job
  DATABASE:  onlinejobs (Jobseekers:$db)   QUERIES: 13 (0.0171 seconds)  (Hide)
0.0004   SELECT *
                                
FROM exrates
                                WHERE rate_name 
'USD-PHP' 
0.0005   SELECT *
FROM `employer_jobs`
WHERE `job_id` = 1703419
 LIMIT 1 
0.0008   SELECT *
FROM `employers`
WHERE `employer_id` = 486653
 LIMIT 1 
0.0013   SELECT COUNT(*) AS `numrows`
FROM `t_thread` `t`
LEFT JOIN `t_thread_misc` `miscON `t`.`id` = `misc`.`thread_id`
WHERE `t`.`job_id` = 1703419
AND `misc`.`idIS NULL 
0.0005   SELECT e.business_namee.logoe.websitee.rebill_datee.date_added member_datehitsDATEDIFF('2026-08-10',ej.date_added) duration_daysDATEDIFF('2026-08-10',e.rebill_date) duration_rebillej.*, e.deactivate FROM employers eemployer_jobs ej WHERE e.employer_id ej.employer_id AND
                                   ((
e.user_level >= '500' AND ej.date_added <= e.rebill_date)
                                   OR 
e.employer_id '' OR (ej.date_approved <> '2000-01-01' and DATEDIFF('2026-08-10',ej.date_added) <= 14 ))
                                   AND 
e.deactivate != AND ej.deleted AND job_id '1703419' 
0.0003   SELECT *
FROM `employer_jobs_skills` `ejs`
LEFT JOIN `skills_categories` `scON `ejs`.`skill_id` = `sc`.`id`
WHERE `job_id` = 1703419 
0.0013   UPDATE employer_jobs SET hit_counts '***Aug-05-2026=1044***Aug-06-2026=359***Aug-07-2026=190***Aug-08-2026=88***Aug-09-2026=2***Aug-10-2026=1' WHERE job_id'1703419'  
0.0007   UPDATE employer_jobs SET monthly_hits '***Aug-2026=1683' WHERE job_id'1703419'  
0.0010   SELECT date_sent FROM jobseeker_sent_emails WHERE jobseeker_id '' AND job_id '1703419' AND status LIKE 'sent%' ORDER BY id DESC  
0.0004   SELECT *
FROM `employer_jobs_skills` `ejs`
LEFT JOIN `skills_categories` `scON `ejs`.`skill_id` = `sc`.`id`
WHERE `job_id` = 1703419 
0.0090   SELECT COUNT(*) AS `numrows`
FROM `employer_jobs`
WHERE `employer_id` = '486653'
AND `date_added` >= '2022-06-08' 
0.0008   select from teasers 
0.0002   SELECT FROM skill_categories WHERE skill_cat_id='' 
  HTTP HEADERS  (Show)
  SESSION DATA  (Show)
  CONFIG VARIABLES  (Show)