Overview

This phase was about **execution**: turning the architecture from Part A into a working, multi-site platform that could support the consulting hub, blogs, docs, and Nouvo Ayiti 2075 without turning deployments into a gamble.

Instead of “one big repo with everything inside,” the ecosystem was rolled out as a constellation of focused projects, each with its own lifecycle but aligned through a shared set of scripts, conventions, and guardrails.

Execution & rollout plan

To avoid breaking the public brand, the work followed a deliberate sequence:

1. **Stabilise local development** – standardise Node versions, scripts, and environment files so every project could be cloned and run the same way.
2. **Stand up separate projects** – create dedicated projects for the main site, blog, and docs with minimal but working content.
3. **Introduce release tagging** – tag meaningful milestones (v1.x, v2.0.0) so the ecosystem itself became a portfolio artifact.
4. **Schedule safe migrations** – only then move domains, DNS, and redirects in carefully monitored steps.

Repositories & pipelines

A key decision was to treat each surface as a **first-class project**, not just a folder:

• **Main consulting site** – Next.js on Vercel, with PowerShell helpers for lint, type-check, and build.
• **Blog site** – static posts generated from Markdown/HTML with an index generator script and GitHub Pages/Vercel hosting.
• **Docs site** – an independent repo tuned for long-form documentation and technical playbooks.
• **Nouvo Ayiti 2075** – its own repo, with multilingual support and a different publishing cadence.

Each repo ships through its own pipeline but follows the same pattern: **pull request checks, preview deployment, manual review, then production promotion**. This keeps the platform consistent without forcing every site into a single monolith.

Routing, DNS & migrations

The most delicate step was aligning URLs with the new architecture without breaking existing links.

• **DNS first-class** – subdomains were mapped intentionally: main brand, blogs, docs, and Nouvo Ayiti 2075 each received a clear home.
• **Gradual cut-over** – DNS changes were performed in small steps, with temporary redirects keeping old links alive while the new routes settled.
• **SEO-aware routing** – important paths like /projects, /join, and brochure links were preserved or redirected to avoid regressions.

The result is routing that feels simple to visitors, even though it spans several independent deployments behind the scenes.

Lessons learned

Executing the ecosystem taught a few durable lessons:

• **Architecture earns its keep in execution** – diagrams are useful, but the real value appears when migrations, pipelines, and DNS all line up in production.
• **Independent projects, shared discipline** – giving each site its own repo and pipeline keeps complexity contained, as long as the surrounding conventions are strong.
• **Document the journey** – tags, changelogs, and case studies turn the platform itself into a live portfolio of engineering practice.

Together with Part A, this execution story shows how a solo engineer can build and operate a small but serious **platform** around their work — one that’s ready to host future tools, courses, and initiatives on top of the current ecosystem.