Automating a Solopreneur Operation: The Systems Behind 100+ Apps
Managing 100+ deployed apps, 45 knowledge domains, and multiple business streams as one person would be impossible without heavy automation. Here's the system.
The Automation Stack
┌─────────────────────────────────────────┐
│ AI Assistant (MCP) │
│ Natural language → Infrastructure │
├────────────┬────────────┬───────────────┤
│ Vercel │ Cloudflare │ GitHub │
│ Deploy │ Workers │ CI/CD │
├────────────┼────────────┼───────────────┤
│ PostgreSQL │ Supabase │ Stripe │
│ Data │ Edge Fn │ Revenue │
├────────────┼────────────┼───────────────┤
│ Analytics │ Uptime │ Error │
│ GA4/CA │ Monitor │ Logs │
└────────────┴────────────┴───────────────┘
1. Deployment Automation
Every project follows the same pattern:
# One command deploys any project
cd project-dir && echo "y" | vercel --prod
But the real automation is in batch operations. When I need to update a shared dependency across 20 projects:
for dir in $(find . -name "package.json" -path "*/book-*/*" -exec dirname {} \;); do
(cd "$dir" && npm update shared-lib && echo "y" | vercel --prod)
done
2. Monitoring & Alerting
Uptime Monitoring
Every project has an uptime check via my Status Page app. If a site goes down:
- Alert fires to Telegram
- Error logs are captured automatically
- AI analyzes the error and suggests a fix
- If it's a simple deploy fix, AI can redeploy via MCP
Analytics Aggregation
Instead of checking 100 Vercel dashboards, I aggregate analytics:
- Cloudflare Analytics — Traffic, bandwidth, threats across all domains
- Google Analytics — User behavior on the main portfolio
- Custom tracking — Per-project events via the Tracking app
3. Content Pipeline
Blog Posts
Idea → Draft (MDX) → Local Preview → Deploy → RSS Feed → Social
The content pipeline is integrated with the portfolio. New blog posts automatically appear in:
- The blog listing page
- RSS feeds (English + Thai)
- The newsletter archive
- Social media previews
Knowledge Base Updates
Each of the 45 domains has curated content. Updates follow:
Research → Article → Domain Page → Cross-links → Deploy
4. AI-Assisted Development
MCP-Powered Workflow
Through the MCP server, I can:
- Deploy any project via conversation
- Check analytics without opening dashboards
- Manage DNS records on Cloudflare
- Query databases for business metrics
- Create GitHub issues for bug tracking
Code Generation
AI assists with:
- Scaffolding new projects (shared template)
- Writing boilerplate (API routes, components)
- Debugging build errors
- Writing tests for critical paths
5. Revenue Automation
Multiple revenue streams, all automated:
| Stream | Platform | Automation | |--------|----------|------------| | Digital Products | Own store | Auto-delivery via email | | Freelance | Multiple platforms | Profile management | | Ad Revenue | Cloudflare/Google | Passive analytics | | Subscriptions | Stripe | Recurring billing |
6. Database & Backup
Automated Backups
# Daily backup script
sqlite3 data/solo-empire.db ".backup '/tmp/backup.db'"
gzip /tmp/backup.db
# Upload to Cloudflare R2
Backups run daily with retention policies. The database stores:
- Project metadata
- Analytics snapshots
- Content drafts
- Business records
7. The Daily Routine
My automated morning briefing:
- System status — All 100+ sites checked
- Analytics summary — Traffic, errors, revenue
- Deployment log — What deployed in the last 24h
- Content pipeline — Drafts in progress, scheduled posts
- Action items — What needs human attention
Key Principles
- Automate the repeatable — If you do it twice, script it
- Monitor the critical — Alert on what matters, ignore noise
- AI for the complex — Let AI handle analysis and suggestions
- Human for the creative — Focus energy on strategy and content
The Numbers
- 100+ deployed applications
- 45 knowledge domains maintained
- ~2 hours/day actual hands-on work
- ~8 hours automated work while I sleep
- 0 full-time employees
The solopreneur's leverage isn't working harder — it's building systems that work while you don't. Automation turns one person into a team.