Already tried ChatGPT. Bought a few tools. Nothing stuck. Watching competitors close faster and feeling the gap widen. Skeptical but open.

A 'in the bag' deal didn't close because the team took 4 days to follow up. Or Q1 review showed $80K in uncaptured leads.

Deals fall through because follow-up is too slow. AI tools bought, never used.

Case study with a real number. The free AI Revenue Audit as proof of how Justin thinks. Building-in-public agent stack as proof Justin does this for himself too.

Direct. No jargon without plain-English. Lead with what he lost, not what Justin does. CTA is the audit, not a sales call.

CEO came back from a conference and decided 'we need an AI strategy.' She owns it with no budget and a team at capacity. Burned by 3-month implementations.

CEO sends her a LinkedIn post about AI follow-up and says 'let's do this.'

Needs a result in 60–90 days or funding dies. Worried team will resist.

Methodology (Revenue Recovery First). Implementations in similar ops complexity. Defined phases. References from ops roles.

Process-focused. Acknowledge the constraint before pitching. Specificity beats enthusiasm.

Managing scale that makes individual optimization expensive. Turnover walks training out the door. Inconsistency across locations costs reviews and repeat business.

Corporate conference announces AI initiative. Or two best location managers quit in 90 days.

Inconsistency across locations. Lead follow-up: some locations close 70%, others close 20%.

Franchise-specific case studies. The 5 Star Painting speaking engagement. Per-location ROI math, not enterprise-level.

Franchise operators speak in unit economics. Per location, per head, per quarter. Lead with consistency before automation.

“I was VP at a digital agency. Named clients. Real budget. Safe income.”

“I looked at what the next 10 years looked like and decided to walk away.”

“In under two months, I built an AI workforce that runs my entire business. That is the proof of concept for everything I build for clients.”

Brand intro videos · bio content · origin story hooks · LinkedIn narrative posts

“Your best employee just quit. Or your manager just had a baby. Or your VP got poached.”

“Everything they knew just walked out the door with them.”

“Businesses that survive this are not more talented — they are more systematized. The process lives in the system, not the person.”

Ray persona · franchise · ops content · high-turnover industries

“People describe what I do in one word: sharp.”

“Not efficient. Not experienced. Not affordable. Sharp.”

“I see the problem faster than most, build the solution before the meeting ends, and the system is running before the strategy deck would have been finished.”

Authority content · competitive differentiation · multi-vendor evaluations

“People expect one thing when they see me. They get something else.”

“The look says I do not need your approval. The work says I am here until your business runs.”

“That gap — between expectation and experience — is the thing that makes people trust fast.”

Brand video content · intros that lead with the visual

Default surface — surface-1 with ghost border.

Elevated surface — surface-2 with faint border.

Transparent — outline only.

Sage-bordered — for emphasis.

Lede paragraphs sit just under headings and breathe with extra leading. Always 300-weight Geist for that engineered, almost-italic readability.

Body copy in Geist 400 at 15px is the default reading size across product, marketing, and editorial. Hierarchy is carried by size, leading, and color — never by adding weight.

For inline emphasis, bump weight to 500 only. Never bold. Never italic. The cap is locked.

Use 13/400 for dense UI surfaces — table rows, sidebars, dashboard panels — where vertical rhythm matters more than reading flow.

1,151

18

$53.20

Soft Bjarmi radial in one corner. The ambient signature behind every hero, header, and CTA surface.

background: radial-gradient(circle at 80% 20%, var(--accent-glow), transparent 60%);

Diffuse radial centered. Spotlight effect for the focal element on a dark surface.

background: radial-gradient(circle at 50% 50%, var(--accent-glow), transparent 70%);

The canonical ambient grain. Very fine isotropic particle noise. Use this — or use nothing.

<!-- via SVG filter, see ForgeFilters -->
filter: url(#niflheim-sand);
opacity: 0.05;

Subtle stone-like surface noise. Quiet weight under content blocks. SVG-generated, zero asset weight.

<!-- inline SVG filter, see Slate component -->
background: var(--surface-1);
filter: url(#niflheim-slate);

Brushed metal grain. Linear high-frequency noise — directional, industrial.

<!-- inline SVG filter with feTurbulence + matrix -->
background: var(--surface-2);
filter: url(#niflheim-iron);

Coarse volcanic-grain texture. Heaviest material register. Use sparingly.

<!-- inline SVG filter, coarser turbulence -->
background: var(--surface-1);
filter: url(#niflheim-basalt);

Agent

Ratatoskr (seo-specialist)

Status

running

Last fired

2026-04-19 09:00 UTC

Next run

2026-04-20 09:00 UTC

Cron

0 9 * * *

Build URL

design-system.justinharris.ai

Deploy SHA

abc12def...

Region

iad1 · Washington DC

1. Production deployDEPLOY14:32

   design-system v1.4 · 38s build
2. TypeScript check passed14:18
3. Layer 2 primitives written14:15

   9 components · 1,400 lines
4. Sage doctrine added to INTEGRITY.mdRULE13:51
5. Tabs underline regression spotted13:32

   User feedback during Layer 1 review

1. RRatatoskr firedCRON09:00:02
2. RGSC snapshot complete09:00:34
3. SSaga briefing prepared09:01:12
4. SSlack post failed09:01:45

   rate limit exceeded
5. SRetry queued09:02:01