Data democratization is key to transformation at Omantel

Two years ago, when Ashraf Dahab Kheiri, Senior Strategist at Oman Telecommunications (Omantel), was leading the development of an enterprise data management strategy at his company, he realized what it would take to make Omantel a data-driven organization: democratization of data across the entire organization. But achieving this is much easier said than done.

Omantel is in the midst of implementing a new corporate strategy which focuses on building AI-based collaborative ecosystems, an effort that is foundational to Oman’s 2040 Vision. This includes an ongoing effort to modernize the company’s data architecture.

“We are evolving into a digital technology powerhouse not only to serve our internal aspirations, but also to serve the key tenants of Oman’s 2040 Vision, which is: digitization, innovation, diversification, entrepreneurship and sustainability,” explains Mohamed Abdullatif, General Manager of Customer Lifecycle Management at Omantel, in this video recorded at DTW-Ignite 24.

Who owns the data?

Like many other telcos, Omantel is evaluating new technology such as data mesh (which decentralizes data ownership), data fabric (which automates data sharing in an ecosystem) and data lakehouses (which combine storage and analytics). At the same time, the company is developing a data governance program to establish clear ownership and stewardship of data.

“Governance, done right, has the answers to the complex problems of organization and change,” says Dahab Kheiri, a former co-leader of TM Forum’s Modern Data Architecture Project which aims to help communications service providers (CSPs) transform their approach to data management. “It forces you to ask questions across the organization: Who really owns the data? What comes with that ownership? What level of authority and what kind of decisions does that ownership entitle people to?”

In many companies, IT teams are often viewed as an obstacle or hindrance to accessing data because in a traditional data architecture, they control it. When data is democratized, business owners are responsible for it.

“Ownership is very important, and governance starts with that,” says Dahab Kheiri. “Then everybody knows what they do. The IT guys realize they are just the data custodians: The data is in their custody, but it's not theirs; It belongs to the business unit that made the sale.”

AI as the ‘North Star’

Enabling AI and automation is an important driver of Omantel’s push to modernize data architecture. “This is an AI age, so anything that does not include AI at the core is a gap. You have to put AI as your North Star,” says Dahab Kheiri. “For us to be able to effectively and quickly perform and be able to test out new and innovative AI capabilities…rigid data warehousing models are no longer relevant – they are just not agile enough. A modern data architecture can help by reducing the rigidity – the boxed mentality – to open these platforms up for innovation and allow you very quick, very rapid access to data.”

The data fabric concept in particular supports Omantel's ambitious push into new areas like fintech, e-commerce marketplaces and digital advertising. For example, the company is building a data lake with data pipelines and AI and machine learning capabilities on Amazon Web Services (AWS) to empower its mobile app, portal and e-commerce shop.

"This is not a telecommunication business model anymore," Kheiri says. "Would it make sense to hire an army of 200 people to copy all of that data and try to force it to be somehow mapped to a telco data model?"

A data fabric allows Omantel to maintain autonomy and agility across diverse data platforms, whether on premises or in the cloud, without the need for rigid integration. "Keep everything separate," Kheiri advises. "Let FinTech have its own data platform and the marketplace have its own. It could be on prem or on a cloud – on AWS, Google Cloud Platform or Microsoft Azure – it doesn't matter."

The key is establishing governance processes to enable seamless data sharing and analysis when needed. Kheiri also emphasizes the importance of open-source technology in enabling the data fabric, calling it an “indispensable” element of a modern data architecture.

“I would go open source even if it cost me more than proprietary software, with the sole purpose is that this particular standard is one that everybody understands so it enables collaborating,” he says.

Empowering users for cultural change

Democratization of data requires that CSPs like Omantel completely rethink how data is stored, accessed and governed, and much of this effort focuses on the cultural change that’s required.

“There has never been a lack of technology – it’s not the systems, not the tools, not the technologies. It’s not even the executive support, because ours are very supportive,” says Dahab Kheiri. “But it’s a huge cultural change.”

He points to the development of analytical dashboards for business units or executives as an example. Traditionally, IT teams within telcos build such dashboards over a period of weeks based on a list of requirements provided by the users. In Dahab Kheiri’s experience, these dashboards typically have a very short lifespan.

“After two weeks of the dashboard being live when you check the systems and see how many times it has been used, it loses relevance faster than you can believe,” he says.

The problem is that a dashboard “freezes a particular perspective of looking at a problem in time,” Dahab Kheiri explains. While business users try to describe their requirements generically when requesting new analytical capabilities, they often are really articulating a specific challenge they are having at the moment. So, the dashboard ends up being irrelevant once the user’s immediate problem is resolved or another challenge becomes more important.

The only way to make data truly relevant to business users and executives is to enable them to answer ad-hoc questions in the moment by analyzing their own data, rather than relying on pre-built dashboards, says Dahab Kheiri. It’s also critical to help business users understand the value of data and why they should use it to support decision-making rather than relying on gut instincts or intuition. This, too, represents significant cultural change.

Data quality and governance are key

In addition to governance, data quality is key to democratizing data. “Data quality is a big inhibitor of the success of data democratization initiatives,” Dahab Kheiri says, adding that it is imperative for users to be able to trust the quality of data.

“It’s a big challenge, especially for the data that that is not machine generated” because humans often make mistakes when entering data, he adds.

Omantel is starting small by offering some basic self-service business intelligence (BI) capabilities. “We start by identifying a group of people who are technically savvy and sit in the business – or if they are not, then we train them – and then we push BI tools to them,” Dahab Kheiri explains. “Sometimes they don’t even need specific tools; it’s done on their browser. And then we give them access to limited curated data sets… We give them the capability of creating their own reports.”

He adds: “We keep expanding this, so this is one way of making data democratization a reality.”