Voice, RTL, and EV: pioneering UX at DEWA

Context

DEWA, the Dubai Electricity & Water Authority, is the UAE’s primary energy and water utility, Dubai’s second-largest publicly traded company, and one of the most aggressive smart-city operators in the GCC. By 2020, DEWA had moved well beyond utility service delivery; the strategic mandate was to operate as a smart-living platform for Dubai residents, covering not just electricity and water, but EV infrastructure, energy management, voice interfaces, and ecosystem integrations.

I joined the Smart Platforms innovation unit in December 2020 as Senior UX Lead, with a brief that spanned three concurrent government innovation projects: a first-in-region Arabic Amazon Alexa skill, the UX for Dubai’s first public EV charging network (Green Charger), and the broader DEWA Smart Living app’s evolution toward iOS-pattern parity.

The unit’s remit was unusual. Most government UX work follows the public sector, slow procurement, conservative tooling, late adoption of consumer patterns. DEWA’s Smart Platforms unit was deliberately ahead of it: shipping voice UI before voice was mainstream in the region, building widgets months before iOS standardised them, and treating EV charging as a hybrid hardware-software UX problem rather than a payments transaction. The work I led there reflects that bet.

The work

Three projects ran in parallel through the year, each with different shape and scope.

The DEWA Alexa skill, the GCC’s first Arabic Amazon Alexa skill, and the region’s first implementation of APL 7.1 with full right-to-left support. Bilingual English / Arabic across both voice UI and graphical UI on visual-supported devices (Echo Show). I owned the VUI conversation design, the GUI flows on supported devices, and the integration architecture. Featured by CNN at launch. Praise from DEWA’s Head of Innovation.

Green Charger, Dubai’s first public EV charging network, 400 chargers at the time across the emirate. The UX challenge was hybrid: physical infrastructure (the charger units themselves) plus mobile interface (booking, locating, unlocking) plus operational backend (DEWA’s internal management dashboards). I designed the end-to-end flow: pre-booking to avoid arriving at occupied chargers, geofence-based unlock when the user reached the station, in-app charger UI within the DEWA Smart Living app, the EV billing layout users received at month-end, and the internal dashboards the operations team used to monitor performance.

DEWA Smart Living app evolution, supporting the broader app’s product roadmap as it matured. This included identifying new iOS interaction patterns four months before iOS released them publicly, building widgets for the app before widgets became a standard pattern in the region, and pushing for a smartwatch app companion. The role was to scout emerging patterns and bring them into DEWA’s roadmap before competitors had to react to them.

My role

Senior UX Lead within DEWA’s Smart Platforms unit, reporting to senior leadership inside the innovation function. The role spanned three concurrent project workstreams; combined budgets exceeded $2M.

On the Alexa skill: I owned end-to-end UX. Voice UI conversation design, every dialogue branch, error path, repair flow, and confirmation pattern in both English and Arabic. Graphical UI on visual-supported devices (Echo Show, Echo Hub) using Amazon’s Alexa Presentation Language. I learned APL 7.1 directly from the Amazon team alongside DEWA’s developers, ahead of widespread documentation availability for the RTL features. Worked closely with engineering and project management through launch.

On Green Charger: I owned the hybrid hardware-software UX, pre-booking flow, geofencing unlock interaction, on-charger UI, in-app charger experience inside the DEWA Smart Living app, the monthly EV bill design users received, and the internal operations dashboards. I conducted user research with charger users, accompanied QA on physical site debugging, and validated charger connectivity performance in the field.

On the Smart Living app: I scouted emerging patterns for the unit, bringing in new iOS interaction patterns ahead of public release, designing widgets before they became standard in the region, and pushing the case for a smartwatch companion app.

Across all three, the unit existed to ship things before anyone else in the region did. My job was to make sure the UX held up when they shipped: researched, defensible, and on time.

Approach

The unifying approach across the three projects was to treat innovation work as research-led product strategy, not as concept demos. The unit had history of building impressive prototypes that didn’t translate into shipped product. I pushed in the other direction: nothing leaves the design phase without a defensible path to launch and a real user research base behind it.

On the Alexa skill, the strategic call was hybrid device strategy. When DEWA initially scoped the project, the obvious play was a voice-only skill for Echo Dot, the dominant Alexa device in the UAE at the time. I pushed for a hybrid: voice-first on Echo Dot, voice + visual on Echo Show. The reasoning was that DEWA’s services (energy bills, water consumption, EV charging status) carried numbers and breakdowns users would want to see, not just hear. The hybrid approach required learning APL 7.1 with full RTL support, at the time an emerging Amazon capability without widespread implementation precedent. I took on learning it directly with the development team from Amazon’s engineers. The bet was correct: the visual-supported flows became the most-used part of the skill at launch.

On Green Charger, the strategic call was hybrid hardware-software UX. Most EV charging UX in 2020 treated the charger as a transactional terminal, tap card, plug in, drive away. I argued for treating it as a continuous service journey: the user’s interaction starts before they arrive (booking) and continues after they leave (billing). Each stage of the journey (discovery, booking, geofence-unlock, charging, billing) got designed as a connected experience. The user research bore this out: the friction in EV charging adoption wasn’t the charger UX, it was the uncertainty around availability before arrival.

On the Smart Living app, the strategic call was pattern-leadership over pattern-following. Most government utility apps adopt new platform patterns 12 to 18 months after consumer apps prove them out. I argued the opposite: DEWA had reach (millions of UAE residents) and could productively shape user expectations of how a government utility app should feel. We shipped iOS-pattern equivalents months before iOS had standardised them, widgets before widgets became expected, and pushed for a smartwatch companion that, even if it didn’t ship in my tenure, set the unit’s roadmap orientation forward, not backward.

Key decisions

01 — Hybrid voice + visual UI for the Alexa skill, ahead of regional precedent

DEWA’s initial scope was Echo Dot only, voice-first. I pushed for hybrid device support, voice on Dot, voice + visual on Echo Show, because the data DEWA’s services surface (energy consumption, water usage, EV charging status, billing) is fundamentally tabular. Voice alone would have under-served the most-asked queries. The decision required taking on APL 7.1 with full right-to-left support at a moment when Amazon’s documentation for RTL was still emerging. I learned it with the development team directly from Amazon’s engineers. The skill launched with hybrid support, became the GCC’s first Arabic Alexa skill, the region’s first APL RTL implementation, and was featured by CNN at launch.

02 — Treat EV charging as continuous service journey, not point-of-sale

The default UX model for public EV charging in 2020 was transactional: arrive, tap, charge, leave. I argued for designing Green Charger as a continuous service journey, pre-booking to avoid occupied chargers, geofence-based unlock on arrival, in-app charging status, monthly billing as part of the same experience. The decision changed what got built (a booking system, geofencing infrastructure, integrated billing UX) and how user research was scoped (longitudinal use studies rather than session-based usability tests). It also changed what went into the DEWA Smart Living app, the EV experience became a core surface within the broader app, not a standalone tool.

03 — Lead the platform pattern cycle rather than follow it

Government utility apps typically adopt new platform interaction patterns 12 to 18 months after consumer apps validate them. I argued, and the unit backed me, that DEWA was at sufficient scale to do the opposite: lead the pattern cycle, not follow it. We shipped iOS interaction-pattern equivalents four months ahead of iOS public release, widgets ahead of regional standard adoption, and pushed for smartwatch parity before smartwatch apps were a Dubai-resident expectation. The risk was real: leading patterns sometimes means shipping patterns users don’t yet recognise. The mitigation was research, every pattern call had a research base showing user readiness before we committed.

Outcome

Alexa skill: Launched as the GCC’s first Arabic Amazon Alexa skill. Region’s first implementation of APL 7.1 with full right-to-left support. Featured by CNN at launch. Praise from DEWA’s Head of Innovation. Bilingual coverage (English / Arabic) across both voice UI and graphical UI.

Green Charger: Public EV charging network operational across Dubai with 400 chargers at the time of my tenure (the network has expanded substantially since). Hybrid pre-booking + geofence-unlock + in-app + billing experience shipped end-to-end. The pattern set the operational template for how DEWA continues to ship hardware-software hybrid services.

DEWA Smart Living app: iOS-pattern early adoption shipped successfully, the app maintained pattern parity with consumer apps months ahead of typical government utility timelines. Widgets shipped before they became regional standard. Smartwatch companion entered the roadmap.

The qualitative outcome: the unit ended my tenure with three shipped innovation projects that other government utilities in the region would reference and replicate over the following years. DEWA Smart Living became one of the more-respected government utility apps in the GCC. The Alexa skill, beyond the launch coverage, became a reference implementation that other Arabic-speaking markets pointed to when scoping their own voice UI work.

Worth being straight about the timescale: utility innovation compounds slowly. The biggest payoffs from the 2020 patterns only became visible over the next 24 to 36 months as adoption matured. I owned one year of that arc.

What I’d do differently

The strongest thing about the role at DEWA, being out ahead of the platform cycle, was also the most expensive habit. I spent significant energy on pattern-leadership decisions whose payoff was 18+ months out, when the unit could have used more of that energy on closing the loop on what we had already shipped.

Specifically: I’d have invested more in measurement and post-launch instrumentation. The Alexa skill launched, got featured, and was strong on day-one usage. I don’t have the post-launch data I should have to answer the question, which queries did users actually rely on, in which language, on which device, and how did that change over months three through twelve? The same is true for Green Charger: we shipped a strong booking-first experience but the longitudinal study of how charging behaviour evolved as the network expanded never got commissioned during my tenure. The iOS-pattern early-adoption work suffered the same gap.

The fix isn’t doing less innovation, it’s pairing every shipped innovation with a six-month measurement plan that tracks not just usage but the right usage shape. Innovation labs default to celebrating launches. The mature version pairs every launch with the measurement infrastructure that will tell you, a year out, whether the bet paid.

If I were to lead a similar government innovation unit again, that’s the single change I’d make from week one.

Artifacts