Omar is obsessed with the idea of ownership. He doesn’t mean vague abstractions of responsibility; he’s talking about the nuts-and-bolts of who owns what – and what that ownership entails. Ownership is sexy in theory, but it’s messy in practice. Everyone wants to ‘own’ something until they realize what that means.

For Omar, ownership begins with expectations. When you own a piece of the system – a pipeline, a dataset, a KPI – it’s not just about saying, “That’s my problem.” It’s about the gritty details of accountability. “If you own a data product, the metadata has to be filled out. The glossary has to be completed. The data quality has to meet expectations, and if something upstream breaks, you better be the one who communicates it.”

He pauses, reflecting on an especially thorny project. “The hardest part is aligning what ownership even means. For instance, does a team own the revenue data as a whole, no matter where it comes from? Or do they only own the system that processes it? These are not small decisions. They affect every handoff in your data ecosystem.”

Yet there’s no such thing as ownership without governance. It’s a word that makes people wince – bringing up images of grim checklists and endless documentation requirements. But Omar insists it’s unavoidable. “Governance isn’t compliance,” he says pointedly. “It’s what makes the whole thing work.”

He breaks it down simply: governance is knowing, at any point in time, where your data came from, how it was processed, and whether you can trust it. “Imagine you’re looking at a report about energy consumption. If you don’t know whether that number came from grid sensors, customer surveys, or a simulation model, how can you base decisions on it?”

“If your end users have to tell you there’s a problem with the data, you’ve failed,” he asserts. “You need to catch issues first. Governance is not just about compliance. It’s transparency. Every consumer of a data product should know its current health, the quality of the data, and who’s responsible.”

Omar doesn’t advocate for perfection. On the contrary, he is skeptical of many idealized models and frameworks that dominate discussions about data infrastructure. The data mesh, he notes, is a concept few organizations actually achieve. “The goal isn’t perfection,” he emphasizes, “and most organizations don’t need it.”

Take, for instance, the decision to build a data platform in the first place. Not every company requires one. “If you’ve got one or two source systems, don’t overcomplicate things,” he advises. “Set up a cron job, drop some files, and let teams work with them. Only start thinking about a platform when you’ve got multiple use cases, multiple sources, and you’re joining data every day with real business consequences.”

For Omar, the goal isn’t a perfect system. “The highest level of data maturity isn’t about some dreamy, centralized architecture,” he argues. “Sometimes the best result looks distributed – teams taking ownership in different ways, adapting to their specific needs.” Energy organizations, he notes, are riddled with inherent complexity.

Amid all the complexity, Omar insists that value remains the North Star. “When leadership starts questioning why they’re spending millions on a platform, what do you say?” The question is rhetorical. Omar has been in that room. He knows what that silence feels like.

His strategy? Tie data use cases directly to KPIs that leadership cares about. It’s something far too many teams overlook. “At one energy client, we mapped out their use cases. Some weren’t adding real value at all – they were just busywork. But a few were core: turbine uptime, grid efficiency, energy consumption forecasts. These are the things that mattered.”

The key, Omar believes, is in prioritization. “The energy sector doesn’t get the luxury of trial and error. If a prediction model for grid demand is unreliable, you risk both financial losses and operational instability. At that point, clarity isn’t a add-on – it’s survival.”

Underlying everything Omar says is a belief in scalability – not just of systems but of culture. As platforms mature, decisions about ownership, governance, and structure define their utility. At scale, the work becomes almost entirely about the people behind the data.

In industries as critical as energy, the stakes couldn’t be higher. “You can engineer the perfect pipeline,” he concludes, “but without the right ownership and governance, it’s a pipeline to nowhere. Scale’s the real test of any system – because once you hit it, solutions don’t matter as much as strategy.” The hum of a data infrastructure may not grab headlines like a turbine spinning at full pace, but for Omar, it’s the unnoticed lifeblood of an industry that quite literally keeps the lights on.