name: ai-operated-business-building
description: Build AI-operated side hustles and online businesses, especially high-ticket B2B service/productized-service models in the user's existing sectors.
version: 1.0.0
AI-Operated Business Building
Use this skill when the user wants to create, validate, or operate a side hustle, online business, productized service, agency offer, AI operator business, or revenue machine.
Local references
references/atlas-growth-desk.md — session-specific notes for the Atlas Growth Desk / 30 Dagen Lead Engine Sprint concept, pricing, revenue math, boundaries, and immediate next actions.references/hourly-builder-sprint.md — operating pattern for hourly autonomous build jobs and manual “ronde” sessions where Seyed wants maximum tangible output.
Default strategy for Seyed
Prefer businesses inside existing advantage zones before random internet opportunities:
- Dubai / international real estate
- Dutch Realtor / investor education and lead generation
- recruitment and staffing
- sales enablement and sales coaching
- AI automation, content systems, and agent operations
Goal framing: build toward €30k-€50k/month with legitimate, executable models. Favor high-ticket B2B first because it reaches revenue targets faster than low-ticket consumer products.
Operating principles
- Do not scatter across random side hustles. Pick one wedge, sell, deliver, then productize.
- Start with cashflow. Use service or productized-service offers at €1k-€5k/month before building scalable assets.
- Exploit existing credibility. Use sector context, network, owned data, prior assets, and operational capabilities.
- AI operates, human approves. Drafts, research, lists, CRM, scripts, assets, reports can be automated. External messages, public posts, paid ads, contracts, and client commitments need user approval unless explicitly delegated.
- Price for margin. Avoid underpriced custom work. Package clear deliverables, boundaries, and renewal terms.
- Measure revenue system health. Track leads sourced, conversations started, calls booked, offers sent, closes, MRR, churn risk, and delivery hours.
Recommended first wedge: AI Growth Ops
Working concept: Atlas Growth Desk.
Sell a done-for-you AI lead generation, content, outreach, follow-up, CRM, and reporting system for real estate, recruitment, and sales teams.
Offer shape
Promise:
In 30 days we build and run your AI sales engine: positioning, content, outreach, follow-up, CRM, and weekly lead reporting.
Initial pilot:
- Setup: €1,500-€3,000
- Retainer after pilot: €2,500-€5,000/month
- Optional performance bonus for qualified meetings or closed deals
Deliverables:
- ICP and positioning audit
- prospect list
- outreach scripts and follow-up sequences
- 30-day content calendar
- 12-30 posts or scripts
- lead magnet concepts
- CRM pipeline
- weekly report
- sales call prep and objection handling
Boundaries:
- no revenue guarantee unless explicitly structured as performance deal
- no unlimited revisions
- no external outreach without approved scripts and target list
- no client claims that cannot be substantiated
30-50k/month math
Fastest path examples:
- 10 clients at €3k/month = €30k MRR
- 15 clients at €3k/month = €45k MRR
- 5 clients at €5k/month plus 10 smaller clients = €40k-€50k/month
Execution sequence
Continuous builder mode
When Seyed asks to keep building this business repeatedly or says to do a “ronde”, switch from planning to asset production immediately:
- Use
/home/admin/atlas-growth-desk/ as the default project path unless he specifies another. - Work in files under the project, not only in chat.
- Produce as many complete launch assets as possible in one pass.
- If scheduling is requested, create an hourly cron job with a self-contained prompt and the external-action guardrails below.
- Report in Dutch with concrete file paths and the next best step.
See references/hourly-builder-sprint.md for the exact cron/manual sprint pattern.
Day 1: Sales assets
Create:
- offer one-pager
- landing page copy
- DM pitch
- email pitch
- intake form
- CRM board
- 3 sample audits or mock case studies
Day 2: Prospect universe
Build:
- 100 real estate prospects
- 100 recruitment/staffing prospects
- 50 warm-network candidates
- fields: company, decision maker, niche, location, channel, reason-to-buy, personalized opener, status
Days 3-5: Outreach and content
Produce:
- approved message sequence
- LinkedIn/Instagram/email variants
- 20 owned-channel posts
- 3 lead magnets
- call booking workflow
Days 6-7: Calls and closing
Prepare:
- audit deck per prospect
- sales script
- objection handling
- pilot close terms
- onboarding checklist
First-response pattern
When the user asks for a business/side-hustle plan:
- Pull local/contextual knowledge if available.
- Recommend a single wedge, not a menu of random ideas.
- Show revenue math.
- Provide a concrete offer and week-1 execution plan.
- Ask for only the access/approval needed to start, or proceed with internal assets immediately.
Pitfalls
- Do not imply external outreach has been sent if only drafts were created.
- Do not ask broad strategy questions when an obvious default exists. Pick a reasonable wedge and move.
- Do not overbuild software before validating willingness to pay.
- Do not use private/client data in public assets without approval.
- Do not make guarantees around investment returns, employment outcomes, or regulated financial advice.