AIUNITES runs 18 connected web applications on middleware we designed and built. Cross-site identity, cloud data sync, open notation protocols, and a shared design system — all in production. That's not a slide deck. That's the portfolio.
Built for how publishing works today. Designed for how it will work when autonomous AI agents arrive.
The AIUNITES auto-publish pipeline runs a 22-site network with no manual deployment steps. Every 10 minutes it commits changes, pushes to GitHub Pages, scans for security issues, audits AdSense compliance, submits URLs to search engines, generates stats, and backs up scripts. One PowerShell script. No server. No DevOps hire.
That's useful today. What makes it worth thinking about for the future is how it's structured: file-based triggers, a queue-based task runner, clear log output, and git as the deploy mechanism. That's the same interface an AI agent uses. Not coincidentally — we built it alongside AI, so AI can already use it.
Auto-publish every 10 minutes · Security scanning across all repos · AdSense compliance checking · IndexNow URL submission · SEO audit and auto-fix · Network stats generation · Automated script backup · GitHub Pages deploy verification · Trademark monitoring · Weekly SEO digest
The pipeline already works the way AI agents work: read files, write files, run scripts, check logs, repeat. An agent that can operate a filesystem and run PowerShell can plug into this today. The script runner has a queue. The admin dashboard shows status. The logs are structured. The infrastructure doesn't need to change — the agent just needs to be capable enough to use it.
Google Search Console API pulling data for all 18 properties · Looker Studio dashboard for combined network reporting · Custom admin dashboard aggregating GSC, AdSense, uptime, and publish stats · BigQuery-ready data pipeline for clients who need warehouse-scale analytics · Automated weekly SEO digests with actionable fix queues
The honest framing: Most businesses aren't ready for autonomous AI agents operating their websites. But the ones who build AI-compatible infrastructure now will have a significant advantage when those agents arrive. We've already built that infrastructure and run it in production. We can build it for you.
Three areas where we have real, demonstrable experience — not theory.
We build the layers that connect systems that weren't designed to talk to each other. Shared identity across applications, data sync between local and cloud, navigation context across sites, and open data formats that move between platforms without conversion.
Vanilla JavaScript, dark themes, CSS variables, responsive design, SQLite databases. We build web apps that are fast, accessible, and maintainable. Our DemoTemplate goes from zero to a fully authenticated app with admin panel in one clone.
We integrate AI capabilities into working applications — not prototypes. Educational games with AI content generation, quiz systems, agent platforms organized by profession. AI as a feature inside a real product, not a standalone demo.
Most consulting firms show you case studies from other clients. We show you the thing we built and run every day. Every site in the webring bar at the top of this page is a live AIUNITES application running on our middleware.
That means when we say we can build cross-site authentication, we can point to 4 apps sharing a user table right now. When we say we can design a data sync layer, we can show you CloudDB deployed to all 18 sites. When we say we can create open data formats, we can hand you the MNN specification — a formal notation system with EBNF grammar, prior art analysis, and a working builder app.
We don't consult on things we haven't built ourselves.
Real infrastructure, in production, across 18 sites.
Shared SQLite user table, session-based auth, demo mode, auto-admin first user, user dropdown with settings, backup, cache viewer, and admin panel. Templated for instant deployment.
Local-first data storage with cloud sync via Google Forms and Apps Script. Online/offline toggle, JSON backup/restore, deployed to all 18 sites. Your device is the source of truth, not the cloud.
MNN (Muscular Neuro Notation) — a formal specification for human movement data that works across exercise, virtual worlds, and robotic control. VRN (Voice Resonance Notation) for vocal training. Plain text, machine-parseable, no vendor lock-in.
A base template with complete auth, admin panel, modals, dark theme, responsive design, and webring integration. New sites go from zero to production-ready in one clone. Consistent UX across the entire network.
VideoBate (fallacy quiz game with leaderboards), Gameatica (math & ELA learning games), AIByJob (AI agents by profession), AIZines (digital magazine creation). AI woven into the product, not bolted on.
UATEST_PLAN.md files for all 18 sites. Feature tracking across the network. Semantic versioning. Systematic test scenarios for auth, modals, and site-specific features. Nothing ships without a test plan.
AIUNITES LLC also operates specialized consulting through dedicated brands.
ERP consulting and Jenzabar services for higher education. Student Information System development, complex T-SQL audit scripts, FERPA compliance, IPEDS reporting, database security analysis.
SIS database audit tools. Production-ready audit scripts for enrollment validation, academic standing, referential integrity, and compliance checks. The AIUNITES network is the dogfooding case study.
The ERP Magazine. Industry news, tutorials, technical deep dives. Bridges database expertise with educational content. Articles about audit scripts, compliance, and SIS best practices.
What we work with. No bloat, no unnecessary dependencies.
| Layer | Technology | Why |
|---|---|---|
| Frontend | Vanilla JavaScript | No framework churn, no build steps, ships fast |
| Styling | CSS variables, custom properties | Consistent theming, dark mode, easy to maintain |
| Database | SQLite | Portable, zero-config, shared across sites |
| Client Storage | localStorage | Local-first, works offline, user owns their data |
| Cloud Sync | Google Forms + Apps Script | Free tier, reliable, no server to manage |
| Hosting | GitHub Pages | Free, auto-deploy on push, HTTPS by default |
| Data Formats | JSON, MNN, VRN (plain text) | Open, portable, no proprietary readers |
| Professional | SQL Server, SSIS, SSRS, PowerShell, Power Platform | Enterprise SIS/ERP development (via ERPise) |
We built and run a 22-site autonomous publishing network. Here's what it costs to have us set one up for you.
The system runs on ~$20/month in tools (Claude Pro subscription). What you're paying for is the setup, configuration, pipeline build, and ongoing support — not the infrastructure. Think of it like hiring an electrician: the wire is cheap, the expertise isn't.
| Service | What's Included | Price |
|---|---|---|
| On-Premises Setup | Full pipeline installation on your Windows machine: auto-publish.ps1, launcher.exe, Task Scheduler, GitHub Pages, MCP filesystem, SEO audit + fix scripts, admin dashboard, custom domains. Includes one 4-hour onboarding session. | $2,000 one-time |
| Network Setup per additional site |
Each additional site connected to the pipeline: repo bootstrap, GitHub Pages, custom domain, webring integration, SEO baseline, admin panel. | $250 per site |
| Monthly Retainer | Weekly Claude sessions on your top action list items, pipeline maintenance, SEO fix reviews, new page builds, visibility gate management. Async via chat. | $1,500 /month |
| Cloud SaaS Workspace | Fully managed cloud version: dedicated Claude instance, cloud filesystem, managed auto-publish cron, multi-site dashboard, team seats. No local install required. | Contact us custom quote |
| White-Label Licence | Full source code transfer: all PowerShell scripts, launcher.exe source, admin panel, pipeline architecture. Rebrand and resell as your own product. One-time transfer + 90-day support. | Contact us custom quote |
What it replaces: A DevOps engineer ($8k–$15k/mo) + SEO manager ($3k–$5k/mo) + content manager ($3k–$5k/mo). The tools cost $20/month. The setup fee pays for transferring the system and knowledge to run it.
Built for how publishing works today. Designed for how it will work when autonomous AI agents arrive.
The AIUNITES auto-publish pipeline runs a 22-site network with no manual deployment steps. Every 10 minutes it commits changes, pushes to GitHub Pages, scans for security issues, audits AdSense compliance, submits URLs to search engines, generates stats, and backs up scripts. One PowerShell script. No server. No DevOps hire.
That's useful today. What makes it worth thinking about for the future is how it's structured: file-based triggers, a queue-based task runner, clear log output, and git as the deploy mechanism. That's the same interface an AI agent uses. Not coincidentally — we built it alongside AI, so AI can already use it.
Auto-publish every 10 minutes · Security scanning across all repos · AdSense compliance checking · IndexNow URL submission · SEO audit and auto-fix · Network stats generation · Automated script backup · GitHub Pages deploy verification · Trademark monitoring · Weekly SEO digest
The pipeline already works the way AI agents work: read files, write files, run scripts, check logs, repeat. An agent that can operate a filesystem and run PowerShell can plug into this today. The script runner has a queue. The admin dashboard shows status. The logs are structured. The infrastructure doesn't need to change — the agent just needs to be capable enough to use it.
Google Search Console API pulling data for all 18 properties · Looker Studio (Data Studio) dashboard for combined network reporting · Custom admin dashboard aggregating GSC, AdSense, uptime, and publish stats · BigQuery-ready data pipeline for clients who need warehouse-scale analytics · Automated weekly SEO digests with actionable fix queues
The honest framing: Most businesses aren't ready for autonomous AI agents operating their websites. But the ones who build AI-compatible infrastructure now will have a significant advantage when those agents arrive. We've already built that infrastructure and run it in production. We can build it for you.
Have a project that needs middleware, a web app, or AI integration? We'd like to hear about it.