Larry Contratti · Design Lead — UX & Product Systems

Business problems,
solved with design.

I'm a design lead who takes tangled internal systems at the biggest companies — Amazon, Capital One, and others — and redesigns them into products that unlock massive business value. I lead the teams that build them: I own the research, align product and engineering, set the direction, and stay in the craft through ship. Here's the impact, in numbers — the design decisions and the stories behind them are available if you want to chat.

Request access → 🔒 Full portfolio is private — ask for access
Happy customers include
Amazon Capital One AARP SiriusXM
$650K/yr
black-box tool retired (SCRAM)
$3.0M/yr
labor value unlocked (Feature Registry)
7 → 23
design org grown & led
7
shipped systems, gov & enterprise

Selected work — the problem, the move, the outcome

Case study details available on request
SCRAMAmazon · 2026 · Sr Design Lead
🔒 Case study
4 days 2.5 hrs
Risk-to-mitigation, on a platform users can interrogate

Its opaque scoring made the incumbent its own audit risk. Leading design across engineering, security, and product, I built an explainable model users can interrogate line by line — and a live 3D globe that makes risk readable across a room.

$650K/yr tool replaced ease 2.2 → 4.8 $1B+ worst-case exposure averted
VOCALAmazon · 2026 · Sr Design Lead
🔒 Case study
26 450
Records captured in a single hour — never touching a form

A 30-minute form was strangling voice-of-customer for Amazon Leo's government business. Leading design with engineering, I rebuilt intake so the user never fills a field — they hand off a note, an AI assistant does the rest — then designed the insight layer on top.

users 4 → 57 form time ~30 min → seconds Marines & Army beta
Feature RegistryCapital One · 2024 · Design Lead
🔒 Case study
$3.0M/yr
Annual labor value unlocked

The team wanted better labels; research proved the model itself was broken. I reframed the problem and led design across product and engineering — a content-first architecture and guided first-run that made a single source of truth usable on day one.

20,000 hrs unblocked/yr users 23 → 200+ UMUX 2.3 → 5.1
Designing the Design OrgCapital One · 2024–26 · Design Lead → Sr
🔒 Case study
7 23
An embedded design team, grown and led

I built and led an embedded design team inside the AI Platform org — hiring, mentoring, and setting the operating model that raised design maturity and, in turn, moved platform adoption.

design maturity 2.1 → 4.2 platform adoption 23% → 57%
FedlaneSolo venture · 2026 · Founder & Design Lead
🔒 Case study
97%
Answer accuracy — a strategic advisor, not a search box

Search is what the market built; advice is what users needed. As founder, I designed, built, and shipped Fedlane solo — a strategic advisor that surfaces federal opportunities months before the RFP.

Solo — design + engineering Trust as a product
AI Feature VersioningCapital One · 2024 · Design Lead
🔒 Case study
3.1 8.9
Platform satisfaction (0–10) — nearly tripled

The PRD had one persona; the system had two — producers who ship versions, consumers who must trust them. I led design across product and engineering to build the versioning model and authoring experience that let teams ship, compare, and roll back AI behavior like code.

sharing rate product → system
Associate Task PortalCapital One · 2021 · Lead Product Designer
🔒 Case study
28
Internal platforms reached by the framework

New associates burned 18 months on tribal knowledge before they were productive — a real headcount cost. Leading design, I built a jobs-to-be-done framework to replace the ramp, then made the data case to scale it across the org.

18-month ramp cycle replaced jobs-to-be-done framework
See the thinking, not just the numbers.

Every number here started as a research finding and a design decision. The full case studies — the reframes, the tradeoffs, and the working prototypes — are available privately.

Request access →

How I work

I start where the business hurts — then use research to prove what the real problem is. I lead the team through the messy middle: aligning product and engineering, making the calls, and staying in the craft until it ships. The systems I design don't just test well — they move the numbers the business actually cares about.

01
Research-led
I run the research myself and let it kill my assumptions. The reframe is usually where the value hides.
02
Systems, not screens
I design the model underneath — information architecture, content, and behavior — so the UI has something true to sit on.
03
Lead through ship
I own the direction and align product with engineering, then stay in the pixels until it's live and measured.

Want to see the design decisions behind the numbers?

The full case studies — research, reframes, and live interactive prototypes — are private. Reach out and I'll get you in.

Request access →