Building UX infrastructure at Europe's leading EV platform.

Back to hands-on: the Charge app revamp

My most hands-on product work over this period was the Monta Charge revamp, the driver-facing app for finding, starting, and paying for a charge, done in close partnership with Denise Tan, Product Director, and Johnny Sørensen, VP of Engineering. Everything else in this case study (Hub, Control, the AI infrastructure) I set direction for and let my team own. The Charge app is where I stayed in the work.

The redesign had to reconcile four different ways a driver actually starts a charge: opening a site from the map, tapping a QR sticker on the charger (no account required), tapping an RFID (radio-frequency identification) charge key against the reader, or using a payment terminal directly at the site. Each path re-uses the same pre-authorization and wallet top-up logic underneath, so the driver never sees the seams between them.

650,000+ public chargers, filterable down to the one that actually works

The map redesign had to do more than drop pins. Drivers filter by availability, charge speed, price, and operator, then get directions straight into Google Maps or Apple Maps. Roaming pushes that further: one integration gives drivers access to over a million charge points across Europe and North America, through partners like Plugsurfing, ABRP, Shell Recharge, and Ionity, without ever leaving the Monta app.

Every way a driver already pays, plus a reason to charge smarter

Payment had to match however a driver already prefers to pay, not force a new habit: the Monta Wallet, credit card, Apple Pay, Google Pay, and regionally MobilePay and Vipps in the Nordics and Twint in Switzerland, all built on partnerships with Stripe, Adyen, and Wise. Smart charging layers on top of that: drivers can set a charge point to only start once conditions are met, low electricity price, low CO2, or a high share of renewable energy on the grid, turning a routine top-up into a small climate and cost decision made automatically instead of something they have to think about. 🌱

A public roadmap, and a room full of customers who already knew the answers.

Two initiatives gave us direct signal before we committed engineering resources to the wrong priorities. I helped launch Monta's public roadmap, where customers submit and vote on features directly, feeding straight into the Sense stage of the ADLC (Agentic Development Lifecycle) system below. I also helped stand up a Customer Advisory Board (CAB): a small group of our best customers, meeting on a regular cadence, giving us first access to what would actually make them renew, upsell, or switch to a competitor, and where pricing made sense before we changed it.

The CAB paid for itself as more than an advisory panel. It doubled as our built-in alpha and beta test group, and became a community in its own right: members compared notes with each other on shared challenges as much as they gave us feedback. It's the same customer advisory board model run by DoorDash, Airbnb, Intuit QuickBooks, Adobe, and Mailchimp: get the people who'd feel a bad decision first into the room before it ships.

Four systems. One backbone.

Each system compounds on the others. Together they form a complete capability layer, from individual contributors to the automated product lifecycle.

The infrastructure above was for the whole company. This is what it did to my own team.

Bridge and the Skills Library weren't just an engineering story. My designers and researchers used them to change how the craft itself gets done: running research and synthesizing findings faster, investigating causation and correlation in usage data instead of stopping at the correlation, generating and iterating on mockups, building coded prototypes to pressure-test an idea before it reached engineering, and in a growing number of cases, shipping coded improvements straight to the app and Hub themselves.

I got back into it personally. After years of the job being mostly strategy and people, I started shipping code again: building coded prototypes for the Charge app and Hub, and doing the design system work in the codebase, not just the Figma library. Updating tokens, components, and documentation as code, not as a visual reference engineering had to interpret.

30 agents. 90% of the product lifecycle automated.

Inspired by Booking.com's experimentation culture and built for an AI-native context. Signal detection to production deployment to performance verification: a closed loop where every measurement feeds the next round of signals.

Four AI agents shipped from internal infrastructure.

None would have shipped without Bridge, the Skills Library, and the monitoring system as the foundation. Each agent was a product in its own right, built on the same capability layer the internal team used daily.

Five education programs. Two leaders. One direction.

88% weekly AI usage didn't happen by mandate. It happened through five structured programs, none of which were required.

Restructure the team you have. Borrow the rest.

I joined as the first UX leader at VP level, but I didn't grow headcount to get there. I restructured the team already in place and leaned hard on cross-functional partners to move fast. Monta is a deep-tech company: charging hardware, grid protocols, payment rails. Progress meant working through the people who already understood that domain (engineers, the CMO, the VP of Customer Experience, and my Engineering peers) rather than routing everything through new design hires.

AI fluency changed what every role meant regardless. I led the rewrite of every role definition, level framework, and JD (job description) across design, research, and content, defining what good looked like in an AI-native team, what player-coaches owned, and how to evaluate people in a world where everyone ships code.

The last thing I built: automated playbooks that shipped after I'd already left.

My final project at Monta was Workflow Automations for Hub: letting any team build automated playbooks triggered by network events, driver actions, or KPI thresholds, no code required. It shipped three months after I moved on, built 100% with AI: design, prototyping, and production code, all run through the same pipeline the rest of this case study describes.

Watching it ship without me in the room was the best proof the infrastructure worked. 🚀

What I'd do differently.

The hardest part of first-in-role leadership isn't building the thing. It's building trust while building the thing. You're being evaluated constantly while operating without precedent. The answer is to lead by example early, delegate fast, and let the team's output speak for the org's value.

On the infrastructure side: governance ships with capability. Always. The AI monitoring system should have launched with the pipeline. The PR quality gates should have been defined before the first designer opened a repo. The Ambassador programme should have preceded the tool rollout, not followed it. Speed created real problems: the designer-to-code pipeline scaled velocity faster than oversight could follow. Shipping capability and governance in parallel isn't a nice-to-have; it's the architecture.