PepsiConnect is how PepsiCo's business customers order and manage their accounts. When something went wrong, too many of them had no way to fix it themselves, so they waited on an agent. Support load was high and the customer experience was slow, and the fastest lever was not more agents. It was letting people solve their own problems.

I designed the support flows for PepsiConnect, including a WhatsApp self-serve channel built for how these customers actually communicate. The goal was deflection that felt like help, not a wall between the customer and a human.

I designed for the channel customers already lived in rather than forcing them into a portal. Meeting them in WhatsApp removed the friction of logging in somewhere new to ask a simple question, and it let the support flow live where their attention already was.

TBD — the specific flow or decision you're proudest of here (how you structured the self-serve paths, what you cut, how you decided what to automate versus route to a human).

Bluebeam's legacy website needed a full redesign, and it had no design system underneath it to build on. The honest way to do this was sequential: build the system, then build the site on top of it. That was also the slow way, and the timeline did not allow for slow.

I led the redesign as Design Manager, and I owned both problems at once: the new site and the design system it would stand on. I managed a small team of designers through it.

I made the call to build bluprnt, the design system, and the new site concurrently instead of one after the other. It was the riskier sequencing, because the foundation was being poured while the house went up on top of it. But building them together meant every component was validated against a real page the moment it was made, and it was the only path that fit the three-month window.

Spectrum customers who needed to reschedule an appointment had to call in to do it. The customer verbatims were clear about the friction, and the fix touched more than one team: onboarding owned part of the flow, tech support owned another, and the change had to work across both without breaking either.

I designed the in-app appointment rescheduling experience, driven by what customers were actually saying. Most of the work was not the interface. It was solving the experience across constraints owned by teams other than mine.

I let the customer verbatims set the priority instead of internal assumptions about what mattered. Rescheduling kept surfacing as a pain point, so that is where I focused, and I designed the flow to satisfy onboarding's and tech support's requirements at the same time rather than optimizing for one and fighting the other.

TBD — the specific cross-team constraint you had to resolve, and how you got both teams to a design they'd support.