Staff product designer and AI engineer. Ten years designing AI products, sole design ownership from zero to one.
›ls work
›
01 · CloudZero
Scaling from a single user to thirty enterprise clients.
Cloud cost intelligence for engineering teams.
Role
Founding designer
Timeline
2018–2020 · Seed → Series B
Partner
Carbon Black · first design partner
Scope
Platform 0→1 · design system · research
Reach
Workflow integration.
Real-time cost alerts in Slack and CLI. Anomalies surface where engineers already work.
The domain
Co-designing with infrastructure teams.
Learning complex cloud architecture directly from backend engineers. Designing tools for the immediate technical team.
Tagging
Adapting to user taxonomies.
Accepting raw AWS tags as-is to build adjacent structures with minimal UI interference.
Design system
Zero Design System.
Adapting IBM Carbon for rapid engineering execution.
The practice
Full-lifecycle ownership.
Research, interaction models, and execution from seed through two funding rounds. Scaling the platform to market.
The data layer
Structure over raw metadata.
Infrastructure tags carry no consistent owner or logic.
Raw AWS metadata
i-0a3f9c2b · EC2 no tag
aws:createdby · IAM-User-12938
s3://logs-tmp-93 orphan
rds:PROD / rds:prod mixed
Name: "temp-new-2"
→
Translation layer
Bucket raw metadata — no code change
Map to feature · team · service
Unowned → unallocated pool
→
Resolved ownership
Payments Platform / Prod / Core API
Team Bravo / Staging
Checkout / Prod / Lambda
Unallocated · 3 → claim
The model
Every resource against its own baseline.
A learned ML baseline detects cost anomalies in your environment in real time.
The instrument
Triage where engineers work.
The platform sends cost anomalies in Slack, routed to the relevant team by design, reducing dependency on dashboards.
The resource, opened on the anomaly. Causes ranked by cost correlation — the deploy on top. Acknowledge reports back to the channel.
Signal over noise
Routine drift logs quietly. A departure from baseline pages the on-call in 33 minutes.
Routed by ownership
Every resource carries an owning team. The alert lands in that team's channel, not a company-wide feed.
Triage before the click
Rate against baseline, start time and projected impact travel with the alert.
Likely causes
Changes from the 24 hours before the spike, ranked by cost correlation. The deploy sits two minutes before the inflection.
Scoped to the actionable
My team's changes is the default view. AWS noise sits one tab away.
The loop closes
An acknowledgement here updates the channel. The team sees it is handled.
The outcome
The first FinOps platform powered by ML.
Competitors monitor cloud cost for finance. The platform maps spend to the engineers who create it.
I design systems, not screens.
02 · Cummins · Product, system, process
Accelerating enterprise product cycles.
Synchronizing software design systems with industrial manufacturing cadences.
Role
Senior product designer
Product
Guidanz — engine diagnostics
Scope
Design system · delivery process · one feature end to end
Design-to-ship
3 months → 2 weeks
Linear3 months
SpecReviewHandoffBuildQAShip
Isolated specifications, sequential reviews, and rigid engineering handoffs.
Living2 weeks
Design in context → comments → ship.
02 · Cummins
Systemizing product design.
Aligning software component architecture with industrial manufacturing cadences.
The system. Tokens — color, type, spacing, radius — and the components they generate. A design change is a change build already has the name for.
Design in the running product. Engineers build against the live screen, not a spec — what they see is what ships.
The resubmit cycle, collapsed. Days of PDF round-trips become one working session on a living design.
An ask sized to a yes. Months of meetings across teams, managers, and directors — distilled to one pilot feature.
The pilot that shipped. One feature, run end to end on the new cadence — the cadence stuck.
03 · Concept · Sound landscape
An audio landscape shaped by emotion.
A minimalist audio terminal removing cognitive overload through responsive geometry.
An interactive appliance for psychological rest. Sound stripped of choice to match your mind.
The interface
No titles. No queue. No settings.
The geometry responds to you. There is nothing to operate.
04 · Greenlight · Diffusion research
Old photographs in motion.
Two years of diffusion research. The first thing it made was a gift for my father.
Take 01 / 05Play
Direction
The instrument
Greenlight — A still-to-film console.
Designed, built, and deployed solo. Wan 2.2 on serverless GPUs. Output resolution matches the source frame.
The live app runs the session behind the reel. GPU asleep between runs.
05 · Visionary · Local AI tooling
Simplifying the AI dataset pipeline.
Consolidating the entire curation and captioning pipeline into one offline workspace. Free and open source.
Role
Sole designer & engineer
Scope
Product · UX · built in Claude Code
Platform
macOS · Apple Silicon
Release
Alpha 0.1.0 · open source (MIT)
The alpha · live
Six ways to see the same photographs.
Curation is re-seeing one set until the right cut is obvious. Each arrangement surfaces a different structure.
Arrange by
Structure gathers near-duplicates before you delete. Color and Person come from perceptual hashes and on-device face clusters.
Every dimension, color, duplicate and tag here is Visionary's real output, on 44 images. The native app does this at a hundred thousand — it's on GitHub.
The hand-off
Packaged for wherever you train.
A curated set only matters if it leaves clean. Visionary writes the caption and packages the folder in the format the trainer already reads.
Caption
JoyCaption's own grammar.
Each caption is built from a type, a length and opt-in extras — no prompt engineering. Taggers and VLMs write to the same .txt.
→
Package
The format the trainer already reads.
Export writes .txt sidecars or one metadata.jsonl. Files renumber in grid order, so your arrangement becomes the training order.
→
Ship
This Mac or a GPU box.
Keep it local, zip it, or scp the set straight to the machine that trains. One way, key auth, nothing stored.
The road to alpha
Five things that didn't survive review.
Good tools are subtractive. Every one of these shipped before it was cut or rebuilt.
Under the surface
Built for speed.
It processes 100,000 photos in under a minute on an 8 GB machine. The heavy analysis runs in the background while you browse.
100K+
images
120fps
sustained scrolling
0
dropped frames in 24 s
Memory footprint against an 8 GB budget.
the rest stays free
08 GB total
267 MB · a 750-photo set~1.5 GB · 100,000 images
The grid decodes only what's on screen, hashes read from thumbnails, and duplicate thresholds are computed once — so scrolling is an array lookup. Memory moved 1497 to 1505 MB across 90 seconds at 100,000 images. One file-descriptor leak surfaced at that scale and was closed.
Design philosophy
Less but better. — Dieter Rams
Designed and built end to end in Claude Code — one design review at a time.
Design
One action color. White.
Dark, neutral chrome. Nothing tints how you read a photograph.
Control
Standard idioms. Ultimate authority.
Finder selection, Quick Look, delete to the Trash. Every destructive step is reversible.
Community
Free. Offline. Yours.
MIT, one command, no account, no telemetry. The images never leave your machine.
This is not the negative.
Previously — creative direction, Belladonna Productions (L.I.E. · A Guide to Recognizing Your Saints · Transamerica)campaigns for Dodge & Keds · a decade behind a camerathis site — hand-built in code with an AI pair
