JAlcocerTech E-books

AI Business: The Guide

First things first.

You cant have an AI business without a data first business.

And a data-first business starts earlier than most people think: at the first form a visitor fills in. Who wrote to you, what they asked for, what they agreed to, and what happened next is the first dataset any business owns. If it’s scattered across five tools, the AI comes later and has nothing good to learn from.

What this book is

This is the boilerplate I built for my own one-person tech business, and the lessons from building it. It’s practical: real configuration, real code, real mistakes.

The setup today:

  • Six web properties: a main site, a consulting landing page, an ebooks library (this one), a technical blog, a product site with an in-browser solver app, and a free website audit tool.
  • One inbound lead engine behind all of them: every form, the audit, the newsletter and paid bookings land in one pipeline with consent, double opt-in, retention, alerts and follow-up emails.
  • Almost no running cost: static sites and one Worker on Cloudflare’s free tier; a homelab for backups and email sequences; a paid email provider as the only bill.
flowchart LR
  A["Attract<br/>articles · free audit · ebooks"] --> C["Capture<br/>same-origin forms"]
  C --> K["Consent + confirm<br/>double opt-in"]
  K --> S["Store<br/>KV · homelab mirror"]
  S --> N["Nurture<br/>emails · reports"]
  N --> B["Book<br/>qualifier → paid call"]

How it’s organised

  1. The Cloudflare stack: Pages, Workers, KV, and where R2 and D1 would fit. What each piece does and what I actually use.
  2. One forms backend for every site: how the contact forms work, from the HTML to the stored lead.
  3. Consent and email: double opt-in, unsubscribe, a do-not-send list, retention, and sending through a swappable provider.
  4. The homelab side: a pull-only mirror, phone alerts and a weekly digest, without opening a single port.
  5. Shipping sites that score well: what auditing my own sites taught me (404s, headers, fonts, share images, scheduled posts).
  6. Pragmatism, with an exit: why Cloudflare today, and the rules that keep leaving cheap.
  7. The fully self-hosted version: the same engine on PocketBase behind Traefik, with automatic HTTPS.
  8. Lessons: the short version, as a checklist.

Chapters 1–6 describe what runs today. Chapter 7 is the plan B: the self-hosted version if a client, a price change or plain principle ever requires it.

Who it’s for

  • Solo founders and small teams who want their own lead pipeline instead of renting a CRM, a form builder and a newsletter tool.
  • Developers who build sites for clients and want a repeatable, privacy-first default.
  • Anyone who has felt the pull between “just use the platform” and “host everything yourself” and wants a middle path with a written exit.

Nothing here is legal advice. Where privacy law comes up, it’s how I applied it to my own sites in the EU.