How I Use AI to Run 8 Businesses (Without Burning Out)

The Reality of Running 8 Businesses

People ask me how I run eight businesses. The honest answer is: I don't — not in the traditional sense. I don't sit in eight offices. I don't attend eight sets of weekly meetings. I've built a system where AI and automation handle the repeatable work, and I focus on the decisions that actually matter.

Here's exactly how it works, tool by tool, workflow by workflow.

The Core Stack That Makes It Possible

Everything runs on three pillars:

• Claude API — my thinking partner for content, analysis, code, and strategy
• n8n — the automation backbone that connects everything and runs workflows 24/7
• Supabase — the shared database layer that every venture touches

That's it. No sprawling enterprise software. No team of 50. Just these three, plus Next.js for the frontend and Vercel for deployment.

Morning: Content Generation for the Publisher Network

The publisher network is one of the most automation-heavy parts of the operation. Every morning, an n8n workflow kicks off that:

1. Pulls trending topics and keyword gaps from our tracking database in Supabase
2. Sends structured prompts to the Claude API with specific tone, length, and SEO requirements
3. Receives draft content back, runs it through quality checks
4. Stages it in a review queue where I spend about 20 minutes scanning and approving

What used to take a team of freelancers several days now happens before I've finished my first coffee. I'm not replacing writers for the sake of it — I'm scaling output in a way that was physically impossible before.

The key detail: every prompt includes context from previous articles on that site, brand voice guidelines, and internal linking targets. This isn't generic AI slop. It's structured, intentional content production.

Mid-Morning: Client Reporting at Byter

Byter is our hospitality marketing agency, and clients expect detailed performance reports. The old way was someone spending hours pulling data from Google Analytics, social platforms, and ad accounts, then formatting it into a PDF.

Now an n8n workflow runs weekly that:

• Pulls data from all connected platforms via API
• Stores the raw numbers in Supabase
• Uses Claude API to generate plain-English summaries of what happened, why it matters, and what we're doing next
• Formats everything and sends it via Resend

Clients get better reports, delivered faster, with more actionable insights. My team spends their time on strategy instead of spreadsheets.

Afternoon: Deal Analysis at Red Cardinal

Red Cardinal is the property investment arm, and deal analysis is where AI really shines. When a potential property comes in, I need to assess:

• Purchase price vs. estimated value after refurbishment
• Rental yield projections
• Local market comparables
• Refurb cost estimates

I've built a Claude-powered analysis tool that takes the basic deal parameters and returns a structured breakdown. It pulls comparable data we've stored in Supabase from previous deals and market research, then gives me a risk-adjusted recommendation.

This doesn't replace due diligence. It replaces the first two hours of number-crunching so I can look at more deals and make faster decisions on which ones deserve deeper investigation.

Throughout the Day: Code Assistance for SaaS Products

Bytable, LittleSpoon, LUFU CRM, Stoneforms — each of these SaaS products needs ongoing development. I write code daily, and Claude is embedded in that process.

Specifically:

• Architecture decisions — I describe what I need, and Claude helps me think through the data model, API design, and edge cases before I write a line of code
• Implementation — Claude helps me write TypeScript with strict mode across all projects, keeping patterns consistent
• Debugging — when something breaks at 10pm (and it does), I can paste the error and get straight to the fix
• Code review — before I push, I'll run significant changes past Claude for a second opinion

Because every product runs on the same stack (Next.js + Supabase + Vercel), there's massive knowledge transfer. A pattern I develop for Bytable can be reused in LUFU CRM the same afternoon.

The n8n Workflows That Tie Everything Together

Here's a snapshot of what's running in n8n right now:

• Content pipeline — topic research, drafting, review queue, publishing (publisher network)
• Client reporting — data aggregation, insight generation, email delivery (Byter)
• Lead routing — new enquiries get scored, tagged, and routed to the right venture
• Monitoring — uptime checks, error alerts, and performance tracking across all products
• Data sync — keeping Supabase, HubSpot, and Stripe data consistent across ventures

Each workflow is a living thing. I tweak them constantly. The beauty of n8n is that it's self-hosted, so I have full control and zero per-execution costs.

The Framework You Can Apply

You don't need eight businesses to benefit from this approach. Here's the framework:

1. Identify your repeatable work. What do you do every day or week that follows a predictable pattern? That's your automation target.
2. Pick a central database. Everything should read from and write to one place. For me that's Supabase. For you it might be something else. The point is: one source of truth.
3. Connect with automation. n8n, Make, Zapier — pick one and go deep. Build workflows that move data between your tools without you touching it.
4. Layer in AI for the thinking work. Once the data is flowing, use Claude (or your preferred model) to generate insights, content, and recommendations from that data.
5. Keep the human in the loop. I approve content. I review deal analyses. I make the final call. AI handles the prep work so my decisions are faster and better-informed.

The Bottom Line

I don't have a secret. I have a system. The system runs on AI and automation, and it lets me operate at a scale that would otherwise require a much larger team and a much larger budget.

The tools are available to everyone. The difference is building the workflows, being disciplined about what you automate, and continuously refining the process.

If you're running even one business without automation in 2026, you're working harder than you need to.