Most of what's on this page, nobody asked me to build.

Method partnered with WebPT and GlobalLogic to evolve an early "Gen3" concept into a real product, split into five parallel workstreams — patient, front-office, back-office, clinical, and one more — each owned by a different AD.

Clinical accuracy was non-negotiable, but it couldn't be solved in isolation — three other workstreams were moving in parallel, each with its own AD, and decisions kept crossing those lines.

I owned the clinical workstream end to end — not every decision program-wide, but ensuring my area delivered value while staying aligned across workshops with leadership from every org involved.

Ran 2–3 pivot concepts down to a single POV for clinical, resolved cross-workstream dependencies as they surfaced, and adopted the client's own design system as the shared source of truth.

Boomerang was a 20-person team — founder, CTO, and CRO included — racing to ship faster rather than pausing to get clarity for end users.

Instead of designing what was asked for, I started asking why before building anything. That opened a bigger question: what should even be on the roadmap.

I initiated Warm Referral myself and owned the workflow end to end — how a lead gets understood in the CRM, how a warm path surfaces, and how the ask gets sent.

One early idea — a Chrome extension letting the product work inside a rep's existing tools instead of forcing a new one — became a real revenue driver. Their first big bet, "Mutual Success Plan," wasn't working, so I pushed Warm Referral instead: surfacing trusted paths into accounts from CRM data. That one worked in market.

Sales Motivator is Zoho CRM's gamification layer — leads, calls, and deals feed into targets, games, and leaderboards, on a platform running for 250,000+ businesses worldwide.

The original dashboard was fixed and dark — every point value and rule was hardcoded, with no way to configure around a manager's own incentive structure.

I initiated this redesign myself and was the sole designer on it, from paper wireframes through shipping — coordinating a cross-functional team of 20.

Replaced hardcoded point values with a criteria-based rule builder — module, action, AND/OR conditions — giving managers two ways to compute points: a flat value, or the record's own fields. Games, Targets, and Achievements now read from that same engine.

As a parent, I kept hitting the same wall other parents had — school communication scattered across calls, paper diaries, and notices.

We didn't start by building — we spoke to 50+ schools first to validate the problem, then built on what we heard. Getting them to adopt something new was harder.

I owned it end to end — product and design myself, sourced and managed a 10-person distributed dev team, and acted as founder-led salesperson, talking to hundreds of schools across US, UK, and Indian markets.

The honest one: we didn't succeed selling into India. School administrations move slowly on new tools even when the problem is real — the people deciding weigh their own priorities first, not just whether it solves something real.

J-ESTA is a new digital pre-clearance system for 71 visa-exempt countries, modeled on the US ESTA — proposed directly to Hitachi, GlobalLogic's parent company.

The current process is dated — no shared data across family members, no digital identity once approved, and a passport you still have to carry physically.

I created the app screens Hitachi is evaluating for 2028. The E-Card — the proposal's biggest shift — wasn't a brief I was handed. It was mine.

Proposed the E-Card, a digital ID replacing the physical passport, one shared record for family travel instead of repeat entry, and two color directions: blue for continuity, red for cultural identity.

Every design review at Method catches usability and accessibility issues — eventually. The problem was timing: gaps surfaced after build had already started, costing real engineering time.

Manual review doesn't scale across three time zones shipping in parallel. Whoever reviewed last — or fastest — decided what "good enough" meant that week.

I built a small AI-assisted copilot for the global service team, flagging usability and accessibility issues at the design-file stage — before a line of code gets written.

The team almost treated it as a pass/fail gate. I pushed back — it flags, a person still decides. AI just changed how fast bad design gets caught.

By 2022, BuyerAssist's product had scaled across a 20-person team with no shared design system — every screen was hand-built, patterns rebuilt from scratch each time.

Inconsistent UI, an inefficient design-to-development handoff, and high overhead on both sides — every feature meant re-solving a problem already solved elsewhere.

I identified the design system as the biggest lever the team wasn't pulling — the inconsistency and rework were too costly to leave unaddressed. I initiated it myself, owning the component library, documentation, and governance model end to end.

Started small and iterated rather than a big-bang rollout, treated the system as its own product with documentation and governance, and validated components with designers and developers before scaling.