        
        Open to Senior · Staff · Principal Product Design roles
      

Product design for systems where decisions carry consequences.

I design AI-assisted and data-rich enterprise products where clarity, trust, and reliable system behavior matter. I turn ambiguous, high-stakes workflows into products teams can understand, build, and operate.

The domain changes.
The systems thinking does not.

View selected work Try the live AI interaction

Previously

EnvoyBlackBerryAsurionAmazonCiscoRealtor.comCheggCitrixInterpres Security+ early-stage startups

Selected Work

2019 to 2026

01 Explainable AI decision support for threat & exposure management Interpres Security0→1 · Conversational AI · Data viz 45→27 min to first insight · analyst trust 62→81% Try the working AI interaction 02 Emergency & threat intelligence platform EnvoyWorkplace ops · Real-time systems Resolution time ↓ 20–30% · response visibility ↑ 40–60% Case study 03 Compliance integrity at enterprise scale EnvoyEnterprise SaaS · Policy tooling Audit traceability 45→96% · approval turnaround ↓41% Case study 04 Threat & exposure intelligence platform Interpres SecurityCybersecurity · Design systems Reports generated up 42% · contrast brought to AA and AAA Case study

05 Skills gap & career readiness CheggEducation · Recommendation UX Turning assessment signal into a next-step path Discuss this project

06 Real estate discovery platform Realtor.comConsumer · Search & browse High-intent discovery at consumer scale Discuss this project

Archive · three more projects

            Interactive
          

Play through the AI interaction I designed, including the state where it refuses to answer.

A working recreation inside the Interpres case study: structured reasoning, evidence links, the “AI Uncertain” branch, and the confirmation every sensitive action passes through.

Open →

How I work

Structure before surface

In consequential software, the interaction model is the reasoning model. I design how a system decides and shows its work before I design what it looks like.

I use Claude and Claude Code, Cursor, ChatGPT, and Figma Make to prototype workflow logic, pressure-test edge cases, and bring design closer to engineering.

Prototypes as arguments

I build working prototypes to settle debates that slides can't. Real latency, real edge cases, real refusal states. Then a decision, not another review.

Trade speed for trust

A system that's occasionally impressive but sometimes wrong is worse than one that's consistently legible. I cut capability to protect confidence, on purpose.

What's next

If you're building something complex and consequential, I'd welcome the conversation.

Focused on Senior, Staff, or Principal Product Designer roles on ambitious, systems-level products.

Get in touch LinkedIn