Thirteen years shipping through review gates: Google launch calendars, pharma legal, a regulated utility, a newsroom. Right now at Dow Jones: six teams, one AI-triaged front door, eight event properties folding into one design system. This page is the artifact I’d hand you on day one. It is a risk register, not a reel.
Warren Velázquez · New York
A Design Program Manager is a connector and a problem solver, so instead of adjectives, here are the six risks that actually sink org-level design programs, each with the mitigation I run and the proof it has worked. If you only read one thing on this site, read this.
Every era is the same job in a new arena: more stakeholders, higher stakes, tighter review. The marker is parked on now.
Relay in production at Dow Jones; an agreements system and a resourcing tool built solo. The agreements system turns what people commit to in a working session into a reviewed, signed record. Relay keeps a human review gate on AI triage.
I partner with commerce & acquisition product design teams at Dow Jones and led 4–6 designers on Google Search Ads. Content: creative, content and localization for android.com, tv.google and wearos.google.com. Marked ◐ partial.
GE Healthcare, Merck, Novartis, NYU Langone, the Endo–Mallinckrodt merger at Prophet, and APS, a regulated utility. Regulated-industry delivery and review-gate discipline, not trust-&-safety policy work.
Owned Monaverse through implementation and testing to launch, plus motion, 3D and video vendors; the Infosys partnership on APS; agency-side at Moving Brands and Prophet. Owning a dedicated accessibility program's budget and conformance targets would be new.
I built Relay, triaging requests from 6 teams, and own the brand marketing team's Asana implementation for reporting, capacity and resource management, supporting ~8 events and 6 launches over the past five months.
I lead the design system unifying the event-site builds into one: 6 of 8 core properties migrated, build time cut ~80%, cross-property inconsistencies eliminated. Standards inherited rather than remembered.
Content + localization
Led roadmap planning and oversaw creative, content, site deployment and localization across three properties, including the Android 12 launch, reprioritizing continuously against immovable public dates.
Planned and managed a year-long UI/UX research and design phase, leading 4–6 designers with Google's Search UX Design Director and his team, spanning new ad formats including video. Requirements shifted by design.
Regulated · 28+ engineers
A comprehensive design system after a 3-month UX research phase and an 8-month UX/UI design phase, stewarded into production with Infosys and 28+ engineers, and skeptical utility executives brought inside the process.
Data → exec comms
Managed the asset-production phase of one of GE's largest marketing campaigns: executive-level visualizations, a global website, hundreds of media assets, and supporting data-analysis delivery. High volume, health-regulated review.
Scale from zero
At a $50M+ (a16z) start-up, sourced and built a team of 3D, UI/UX, motion and visual designers, then defined the project plans and production methodology they ran, across 12 metaverse platforms plus web2 properties.
Vendors & contracts
Owned the Monaverse vendor relationship through implementation and platform testing to a single coordinated launch, and managed external vendors across motion, 3D rendering and promotional video.
The JD asks for both halves, and most candidates answer only the numeric one. Every program I run gets one measure you can count and one measure you can only observe, defined at kickoff, not reverse-engineered at review time. Here are six real programs, both halves each.
Thousands of spreadsheet cells of half-finished planning, in six incompatible formats, read, reconciled and resourced by AI so a human never has to. A human review gate before triage becomes work.
It turns what people commit to in a working session into a reviewed, signed record, so a scope or a hand-off cannot quietly evaporate.
The meeting updates the plan, instead of someone updating the plan after the meeting. Nothing changes unless a human confirms it.
The same hour goes wrong in every creative team I have worked in. Agencies, a startup, in-house teams. The resourcing meeting happens, real decisions get made out loud, and then somebody has to remember them and write them up afterwards. Usually late, usually thinner than what was said. The plan drifts out of date until the next meeting rediscovers it.
So I built the tool shown here. The resourcing decision gets made out loud in the meeting. It is captured as it happens and staged for a one-click yes, so nothing changes unless a human confirms it. Once confirmed, the capacity view and the team's plan of record update themselves. Capacity is computed rather than assumed: PTO, standing meetings and partial weeks come off the top before anything is promised, so an over-capacity week shows up before the work is promised away.
These are operations systems, not ventures. Each one was built for a specific problem I ran into in the workplace and designed around that problem rather than around a demo. Each is a real build: several tools integrated so the work does not get re-typed between them, a custom database underneath, AI used where it earns its place, identity and permissions designed in, and hosting and infrastructure I set up and keep running. I specify the system, direct AI agents to build it, review what comes back and send it round again, with staff engineer friends advising. I am not an engineer and I do not hand-write production code. The skill on show is knowing what to ask for and recognising when the answer is structurally wrong.
Status, plainly: it is a fully built tool, in real weekly use, that I built to solve a problem in my own role. The honest limitation is that it is built for and used by me and my own planning, not deployed across my employer. Happy to walk through it live if this turns into a conversation.
The resourcing hour
In the meetingThe meeting updates the plan, instead of someone updating the plan after the meeting.
Collaborations usually start on a handshake, and the terms get reconstructed later, when it's expensive. So I built an agreements system: it turns what people commit to in a working session into a reviewed, signed record, so a scope or a hand-off cannot quietly evaporate. That is consent and provenance as design decisions, made before the work begins rather than argued about after it.
It's evidence of a habit: when the risk is real, I make it a build requirement. Relay keeps a human review gate between AI triage and real work. The resourcing tool stages every decision for a one-click yes, so nothing changes unless a human confirms it. That habit is what a Design Program Manager is actually being hired for. Not the vocabulary of responsible AI, but the instinct to put the gate in before someone asks for it.
I have not run a formal bias-audit program, and I have not worked in trust & safety. What I have done is ship AI products where consent and accuracy were non-negotiable, and integrate an AI tool into a real workflow at a company with real editorial and legal review.
Real references from real programs: a Google director, a designer from the APS build, and an executive director on the agency side.
Warren owned creative, content, deployment and localization on android.com, with meticulous capacity planning to hit our targets across every workstream, without burning the team out. Exceptional communication and problem-solving.
Across a year-long APS program, Warren ran PM and client management on a complex, multi-part project. A genuine team player and problem solver who built strong client relationships throughout.
Exemplary PM on a complex, large-scale project: diligent planning, real adaptability, and expectation management on both the internal and external sides. His role was critical to the project's success.
If your design org needs someone who writes the register, names the gaps, and measures both halves of impact, I’ve been doing that job under other titles for thirteen years.
Why Meta, and why now: Meta is putting AI tooling inside its design org faster than any company its size, and the open question is who keeps the review gates human. That is the exact program I have been running at Dow Jones, at smaller scale, since February.
About the timing. February makes it about six months, and I am not running from anything. The work is real, I am doing well at it, and I would happily keep doing it. What is moving me is a preference that got clear over thirteen years, not over the last six months: I do my best work where design operates at a scale that forces the systems to be real, and where the question of how AI enters a designer’s day is being answered in production instead of debated. That describes very few companies. Meta is one of them, and that is worth acting on when the role is open rather than discovering in month four that I let it pass.
Send me how work reaches your design org today and I’ll tell you where the leaks are.