Fluent by Design
Ship in days. Raise the bar. Prove it moved the business.
I install the AI operating layer your product design team is missing, in 90 days, and I keep it alive every week after. Brief to validated prototype goes from weeks to days, the system transfers to your team, and the gains get measured against your Week 1 baseline.
Free · five taps · the honest read · no signup
Every company has the mandate. Almost no team has the method.
IBM put numbers on the gap: 86% of CEOs believe their people already have the AI skills. About 25% of workers use AI with any regularity. Leadership reads the mandate as handled. Up close, it's a few private experiments, lots of generating, and nothing that compounds.
The result is activity everywhere and outcomes nowhere. Teams are drowning in output with no shared bar for what's worth building and no proof that any of the speed moved a number the business cares about. AI made production cheap. It made judgment and proof scarce. Those two are what this installs.
“My CEO keeps asking what we're doing with AI and I don't have a real answer.”
“We bought Figma Make and Cursor and neither is being used.”
“Engineering is shipping features faster with AI while design looks slower.”
What gets installed: three gates, not more tools
An operating model installed into how your team already works, on whatever tools you already use. Nothing here is bound to Figma or any single platform. The judgment raises the bar. The agnosticism lets it travel.
Gate 1 · Intent
Is this worth starting?
Every piece of work maps to a business goal and a real customer pain before it begins. Governs which work is worth starting, so speed never gets spent on the wrong thing.
Gate 2 · Decision
What do we build, and what does good look like?
A shared, written bar set before work starts. This is where design judgment lives, and the one part AI cannot do for the team.
Gate 3 · Value
Did it move the needle?
Every accelerated ship gets checked against the number it promised. The proof loop, and the reason leadership keeps funding the work.
The machine underneath
The gates run on Design Team OS, an open source library of gate tested skills plus a conductor that routes the loop: which gates are proven, which are open, what can run next. One work ledger per feature carries the evidence across sessions and people. Artifacts, never checkmarks. It installs into a team's workflow in one command as a Claude Code plugin, minutes instead of a setup project.
The proof we sell
Two layers, read straight from the work ledgers the machine keeps.
Layer one · Leverage
The work got faster and cheaper. Time from brief to validated prototype compressed, more directions explored per brief, cost per validated learning down.
Layer two · Outcome
The speed paid off. A real metric moved and mapped back to the business goal, from attitudinal signal at prototype to performance after ship.
Anyone can claim the first layer. Almost no one can draw the line to the second. That line is the product.
What this is not: not another AI tool, your team has plenty. Not an AI tool audit or vendor consolidation exercise. Not design leadership coaching, you coach your team and I install the operating layer around you. Not a design system rebuild, I work inside the one you have. Not a one off workshop. It is the bar and the proof, the two things AI made scarce instead of cheap.
The engagement
Like an advisor, not an agency: the IP and the system transfer to your team, and I stay in the room for 90 days until it runs without me. You are not renting hours. You are buying a working pipeline you own.
Phase 1 · The install and transfer
$10,000 to $15,000one time · scoped to your team · 90 days
- Brief to validated prototype in days instead of weeks, measured against your Week 1 baseline from day one
- A real answer to the CEO's AI question: every initiative tied to a goal and checked against a number
- The stack you already bought finally gets used, with a skill library tuned to your team's real work
- Everything transfers: the gates, the skills, the playbook, the ledger. Your team owns the system when I leave
Phase 2 · Kept alive
from $1,500/month · month to month · after the install
- Every skill in your team's library gets updated the week the tools change, so the playbook never goes stale
- From prototype to MVP: support carrying validated work into the build
- Monthly pulse against the scorecard, quarterly stack refresh
- Async support for the questions between calls
The guarantee
Three validated prototypes shipped through your new AI pipeline in 90 days, measured against your Week 1 baseline, or Phase 2 is free until we get there.
The risk sits on my side of the table.
One install at a time, by hand, for one team. That's the offer, not a limitation.
The math against the alternatives
Built for the leader holding the mandate
Heads of Design, VPs of Design, and Design Directors at product led companies of roughly 50 to 500 people, with design teams of 5 to 20. You have the mandate from above, real motion below, and no line yet between the two. If that's you, the baseline will tell you exactly where the line breaks.
It requires a design team of five or more and a functioning design system. If you're pre seed or pre design system, the open source repo is the right entry, not the install.
How it starts
- 1
Run the baseline
Sixty seconds, five taps, the honest read on where your team actually is. No signup.
- 2
A 30 minute working session
Bring your last shipped prototype or current brief. Leave with the plan mapped onto your actual workflow and the bar drafted.
- 3
A 48 hour sample
Skills tailored to your team, built against your real work, before any contract exists.
- 4
The 90 day install
A proposal in plain language, then the gates go into your workflow. One team at a time, delivered in the room.
Free, before any of the above
The team AI baseline
Sixty seconds to an honest placement on the maturity curve, plus the full written readout and 14 day plan by email.
Fluent by Design, the newsletter
One issue every Friday on the gap between AI mandates and what teams actually ship.
Design Team OS
The open source machine itself. Gate tested skills, the conductor, the work ledger. MIT licensed, installable in one command.
Who's behind it
Fluent by Design is Roy Vergara: ten years of design leadership at Whole Foods and Amazon, retail environments where the teams that survive are the ones with the tightest operating loop. I ship AI products in public and build this system in the open, one skill drop and one Friday issue at a time. Everything sold here runs on the same machine that's free on GitHub. Questions before booking anything: roy@fluentxdesign.com.