eMabler delivers 99.999% EV charger availability with Microsoft Fabric

“In Norway, more than 95% of new cars are EVs,” explains eMabler CEO Juha Stenberg. “In the Nordics, it’s over 50%. If someone buys an internal combustion car now, the reaction is, ‘What’s wrong? Why didn’t you buy electric?’”

In much of the world, EV driving is no longer a novelty. It’s normal. And because it’s normal, the charging experience must be too. As Stenberg puts it, “Charging must become invisible.”

When Stenberg and eMabler CTO Ville Parviainen started the company, the challenge wasn’t demand, as Norway, Finland, and Sweden were electrifying faster than any set of regions outside China. To eMabler’s customer base of charge point operators (CPOs)—the retailers, parking operators, municipalities, and energy companies that own EV charging stations—EV charging was simply an inevitable extension of their digital services.

The eMabler vision: API-first, invisible, and integrated

From its launch, eMabler didn’t set out to build yet another EV charging ecosystem. The company didn’t want to own the user experience, nor were they out to force retailers, utilities, or parking operators to use a white‑label app. Instead, eMabler’s strategy was to disappear into the systems that its customers already relied on. That approach has paid off.

Today, the eMabler charge point management system provides consistent, highly reliable charging services across more than 90,000 chargers in 11 countries, enabling more than 1 million charging sessions per month. The eMabler Connect platform gives customers the ability to manage their EV infrastructure and service as it flexibly adapts to diverse climates, regulatory models, and business models.

“We’ve always been API-first,” Parviainen says. “If a retailer has a loyalty app, we integrate into that. If an energy company already invoices hundreds of thousands of people every month, we feed the EV charging row into that same invoice. We’re not asking customers to change their world to adopt EV charging—we fit into the world they already have.”

That philosophy became eMabler’s key differentiator. Whereas competitors sold vertical, closed stacks—app, CRM, billing, charger management—eMabler sold the opposite: openness.

For retailers and food services operators, activation happens inside loyalty apps. Parviainen notes, “Most customers in the Nordics are already using a retailer’s mobile app, and we enable the addition of EV charging to their existing app. They didn’t build a new app. They just used ours underneath.”

In this way, EV charging becomes a driver of revenue. Parviainen gives a personal example, explaining, “Like most people, when I’m driving with my family, I stop to charge where we can also eat at the same time. It’s a better experience for us, and the retailers who end up making more money from the coffee and lunch than from the electricity.”

For parking operators, eMabler integrates charging with parking billing. “For the largest parking operator in the Nordics,” Stenberg says, “you used to open one app to pay for parking and another to activate charging. Now you pay for both at the same time.”

For utilities, EV charging becomes part of the household energy relationship. ”Since most of the EV charging happens at home, integrating EV charging with energy management is the game changer for energy companies,” says Stenberg. He explains this strategic shift: “Utilities already sell electricity to the customer. EV charging is just another, very natural, extension. And with our platform, they can balance when charging happens, using dynamic pricing as an incentive for people to charge during times of lower demand.”

Transforming at speed

The openness that eMabler offers came with a number of technical challenges. Integrating into everyone else’s systems means your own platform must be flawless. That requires real-time telemetry, accurate billing data, reliability metrics, the ability to standardize events, and ensuring new charger vendors onboard with ease. To assure all those capabilities going forward, the team realized eMabler needed a single, consolidated, infinitely scalable data and analytics engine.

When eMabler Data Engineer K.V. Gopalkrishnan joined eMabler in 2024 to lead the effort, he found an ecosystem that had grown faster than its foundations could realistically support.

There were three separate databases, each with its own purpose but no unified analytical layer: Microsoft Azure Cosmos DB storing operational charger objects, Azure SQL Database storing billing and transactional events, and Azure Data Explorer ingesting real-time Open Charge Point Protocol (OCPP) telemetry. Azure Service Bus and Azure Event Hubs received charger events and dozens of ad hoc scripts and queries gluing things together.

It did the job, but not without a good deal of human effort. “The data was everywhere, across three databases,” Gopalkrishnan recalls. “Some calculations had been attempted, but they weren’t running fast, and you couldn’t put them into production.”

With no pipelines, no modeling, no unified format, and thousands of new chargers coming online, the situation was unsustainable.

Because most of its CPO customers run on Azure, eMabler was a Microsoft shop from the outset. That relationship led Gopalkrishnan to evaluate a range of potential Microsoft solutions to uplevel eMabler’s analytics capabilities: Azure Databricks, Azure Synapse Analytics, Fabric notebooks for raw data, manual ETL bursts, and experimental pipelines. He pieced together prototypes, but nothing felt cohesive.

Speaking with Microsoft about the challenges he was facing, Gopalkrishnan was introduced to Fabric.

The following weekend, he began to familiarize himself with Fabric and immediately felt he was on the right path. “Because I had done so much background work on the data, I was able to put something up on Fabric over a weekend,” Gopalkrishnan recalls. “And once I put it together—I saw it. This was it. This was the platform.”

In short order, he understood that everything that had been scattered—Azure Cosmos DB operational data, SQL transactional data, Azure Data Explorer time-series telemetry, OCPP events—could exist in one place. Not only that, it could be governed, unified, and queryable at real-time speed.

“Fabric simplified the architecture, bringing everything into one place,” Gopalkrishnan says.

Azure Cosmos DB and SQL Database were mirrored directly into Lakehouse in Fabric. Azure Data Explorer telemetry still flowed, but notebooks now performed the heavy analytical lifting without pushing the analytics platform beyond its limits.

Running on Azure and Fabric, eMabler can optimize and scale its business‑critical EV charging workloads intelligently across countries, customer environments, and diverse hardware ecosystems. In this way, it will help ensure consistent 99.999% availability as the network expands.

Fabric didn’t just unify data; it made it possible for eMabler to build products on top of the data. By building its platform on Fabric and Azure, eMabler is creating the data foundation needed to become a true platform for AI innovation in automotive, bringing new intelligence, automation, and reliability to the mobility ecosystem.

Insights that drive growth

The first major Fabric-powered product was eMabler Data Insights, a deep analytical environment available in eMabler’s Connect UI. Parviainen describes its impact, saying, “Using Data Insights, we can show current and prospective customers success rate numbers, even when they don’t have any chargers connected yet, because we have accessible data from tens of thousands of chargers across our customer base.”

For the first time, eMabler could present a CPO evaluating hardware with data on success rates on a range of metrics. These included charger manufacturer X versus Y, firmware issues on specific models, utilization patterns across similar site types, reliability over temperature conditions, and typical failure modes across hardware families. This not only enhanced the company’s consultative role, but also helped transform operators’ decision making. Rather than reacting to unforeseen issues down the road, they could plan proactively and avoid them.

Because Data Insights is built on the unified analytical model of Fabric, rather than static reports or stored dashboards as in most EV platforms, everything is refreshed continuously, with real-time telemetry. As a result, eMabler can tell customers not only what is happening across their charging networks, but also why.

Automating the future

If Data Insights lets customers see, the next product—Pulse—lets them act.

Paired with Data Insights under the “eMabler Sense” umbrella, Pulse is eMabler’s automation engine, designed to fix issues before users ever encounter them. Gopalkrishnan describes it as the start of eMabler’s autonomous operations layer. He anticipates enabling Pulse to perform actions such as automatic remote reboots, firmware updates, configuration corrections, error sequence detection, and preemptive mitigation of known issues, reducing the need for field workers to perform these tasks on site. Going a step further, eMabler is working to empower Pulse with predictive maintenance triggered by real-time anomaly detection.

Parviainen explains where this is going. “We want to specialize models that interpret what the customer support teams used to do—and solve it automatically,” he says. “The dream is that nobody even calls because the issue was solved before anyone noticed.” This aligns with Stenberg’s central philosophy: “When the driver goes to the charger, it needs to work.” Pulse is how eMabler makes that dream possible at scale.

The company’s next objective is creating fully autonomous operations, with AI analyzing charger behavior in real time, detecting anomalies, categorizing failure modes, and triggering automated fixes across thousands of sites. Gopalkrishnan looks forward to the innovations that Fabric will help eMabler achieve as the company targets doubling the number of chargers it supports.

“I’m confident that as we set out to create a full end‑to‑end machine learning, AI, and analytics platform with Fabric to streamline predictive maintenance, enable dynamic pricing and energy optimization, and launch self-healing charger networks, we’re positioning eMabler as a platform for AI innovation in automotive. One that drives smarter, more reliable charging experiences at scale,” Gopalkrishnan concludes.

Discover more about eMabler on LinkedIn.