Built by Logan / Case study / Making things people see

Vista Site and Insight Insights

The site you are reading is the proof. A production business platform with a custom design system, a content engine that researches and drafts articles into a human review queue, its own admin console, analytics, checkout, and membership wiring, built entirely by talking to AI and shipped through 77 pull requests.

AMaking things people see
CSoftware you own
BHard-to-get data, made usable

$170,000+: what a dedicated development team would charge to build it
$25,000: what Logan would charge to build it for you
About$480: in software and AI usage to build it internally (your team's time excluded)
8months: from first commit to a live platform that keeps shipping, built alongside every other project here

Team figure reconciles independent replacement-cost estimates, not a quote. Internal figure is reconstructed from usage records. 349 commits across 36 active build days, 77 merged pull requests.

What problem does it solve

A credible presence usually means an agency, a template, or both.

Agencies charge for a brochure and leave the business with nothing it can change. Templates look like templates. Content costs a ghostwriter per article. Membership means a platform that takes a cut and owns the relationship. A new advisory business needs all four and can afford none of them the usual way.

What impact does it have

One platform. Every funnel. Two articles a day.

The site carries the marketing pages, offer pages, advisor profiles, booking, and every funnel on one custom design system. The Insights engine researches, drafts, fact-checks, and stages articles into a review queue with a searchable catalog, topic filters, author archives, and scheduled publishing. Checkout, webhook-verified membership, community role grants, and launch emails run without a membership platform in the middle.

What value does it have

The whole stack, owned and compounding.

Every page, pipeline, and integration is code the business controls, deployed on infrastructure that scales for free. Each article adds to a search asset the business keeps. Each conversion event feeds an analytics loop the business can read. And the same platform is now hosting the case studies you are scrolling through.

What it does today

A live production platform. Everything below is running on the site you are on.

Site and design system

Production web platform on a modern React framework with server rendering and incremental regeneration
Token-based design system: palette, type, motion primitives, scroll reveals, count-ups
Marketing pages, offer pages, advisor profiles, booking integration, live events index
Structured data on every page for search and answer engines; sitemap, feeds, canonical metadata

Insights content engine

Skill-driven pipeline: research, draft, fact-check, stage
Topic queue, reader-intent taxonomy, search, format and topic filters, author archives
Human review queue with a publish gate; auto-publish deliberately off
Scheduled publishing with future-dated posts; generated hero images per post

Admin and analytics

Unified admin console: content review, newsletter operations, subscriber management
Product analytics through a reverse proxy: funnels, replays, heatmaps, error capture
Conversion event layer wired to every call to action
In-house newsletter with compliant mailing address and unsubscribe handling

Commerce and membership

Pinned-price checkout through a first-party route
Webhook-verified purchases that grant and revoke community roles automatically
Buyer persistence, launch and nurture emails, curriculum distribution through gated links
Gated preview routes for drafts and unreleased pages

On the roadmap

Case-study showcase under the AI Cohort page
Newsletter first send
Interactive demo hosting

Why owning it matters

No agency in the loop.

A new page, a new offer, a new funnel is an afternoon, not a statement of work.

Content is an asset, not an invoice.

Each article compounds search authority the business owns. A ghostwriter's article is spent the day it is paid for.

Membership without the middleman.

Checkout, verification, and community access run on the business's own code. No platform takes a percentage or owns the member list.

Why the operator built it

Why the founder built the site instead of buying one.

The site is the pitch.

A business that teaches operators to build their own software cannot credibly rent its website. Every page had to be evidence that one person talking to AI can ship a real platform, and it is.

Seventy-seven pull requests of taste.

Custom design systems are what agencies charge the most for and templates never deliver. Iterating with AI, pull request by pull request, produced a look the founder chose rather than one a template allowed.

Eight months, always alongside.

The platform grew in the gaps between client builds, one shipped improvement at a time. That cadence, small and constant, is what the Cohort teaches, and this site is the longest-running example of it.

That is what the Vista AI Cohort teaches: operators building their own software, with a community for the parts that need a second set of eyes.