Telefónica modernizes data architecture for autonomous network strategy

Data democratization is a challenge for a large, 100-year old company like Telefónica. Legacy systems have created data replication, duplication and silos, but this is being overcome thanks to new technologies and approaches, explained Paola Saputelli, Global IT Director at Telefónica. In an interview with TM Forum Inform, she shared the operator’s progress towards a modern data architecture. 

“Data democratization and strong data governance are the fundamental pillars for any company. We are working on that because we aspire to be truly data driven, and … the only way to do that is to work on data democratization – and not only the technical part but also as an end-to-end picture of the company,” said Saputelli. 

This is a technological as well as cultural change for telcos. Saputelli said a change in mindset is key to being able to leverage the operator’s “most important data”, the business data.

Telefónica is in the early stages of its efforts to modernize data architecture as part of its Autonomous Network Journey. The operator has defined a strategy for enabling data democratization that involves standardizing common data models across different operating business units and adopting data mesh and data fabric concepts.

“The key is to enable data access without compromising quality and governance…Our approach is based mainly in data virtualization, which allows us to share information between different businesses across the functional domains without the need to replicate,” she explained. 

This approach also helps the company ensure consistency in terms of data quality and efficiency when it comes to storing and accessing it.

OSS data is first use case for data architecture

Telefónica’s data architecture modernization started with OSS data, specifically inventory and assurance, and the company is creating a roadmap to expand to other use cases and domains. Assurance is a strategic use case because all alarms in the network are managed with data and all incidents create alarms. Indeed, Telefónica hopes to reach Level 4 autonomous network (AN) this year for assurance using its modernized data architecture, said Saputelli.

Improving the organization, accessibility and governance of OSS data is a critical first step in the company’s AN strategy, according to Carlos Medina Gonzalez, Head of Mass Market, Channels & Digital Enablers Architecture. This requires first developing a common data model for use across Telefónica’s operating companies and the domains within them, followed by development of the new data architecture. 

“For example, we have to ensure that for customer engagement, for OSS, for finance, the information is the same with a common model…and this is complemented by the implementation of a robust data governance model that ensures data quality and security, which are a very important aspect also and traceable through the whole organization,” said Saputelli. 

Telefónica’s data architecture includes data ingestion and storage capabilities as it always has, but the company is also using data lakehouse technology to combine storage and analytics. Telefónica uses a mix of public and private clouds plus on-premises storage, and the company works with all the big hyperscalers plus other technology partners that provide its telco cloud and data virtualization and visualization capabilities. 

Embracing data mesh and data fabric

Telefónica is adopting a data mesh and data fabric approach to democratize access to and use of data.

The data mesh covers “how we organize data, how we structure the domains, and also what are the responsibilities … to guarantee the quality of data,” Medina explained. This includes looking closely at the roles involved in data governance. “We need data stewards and data owners in order to guarantee the value of this data, the quality of this data,” he said.

It is critical that owners take responsibility for ensuring the quality of data, “because we are not going to replicate this information,” said Medina. “Each domain has a responsibility … to guarantee that this data is the best data that we have, and also that we can share with the rest of the areas of the [organization]. This is a very important thing.”

The data fabric is about “how we interconnect all this data, how we give the flexibility to share this data between all the domains,” Medina explained. Here the concept of data virtualization is important because the resulting data products are information that can be shared in real time. 

The need for industry collaboration and standardization

Saputelli and Medina believe that collaboration on a common data model for the telecoms industry is needed and that the TM Forum’s Information Framework (SID) could be improved to become the model.

Today, Telefónica’s group function has standardized on the Information Framework, but some of the company’s operating businesses use their own data models. This creates extra work because data teams must transform data to fit the Information Framework model when it’s ingested from various domains.  

“I think we need to have a model that we can share between all, and I think one of the enablers that we can have is to push more vendors to use it… If we have the same understanding of the data and the same structure of the information, we can accelerate [data architecture modernization] a lot,” said Medina.

TM Forum should lead the initiative to develop a common data model, said Saputelli, adding that there needs to be a balance between prescriptive standardization and flexibility of implementation.

Medina agreed: “We discuss sometimes with TMF that in some cases the SID and also the APIs are very flexible… I know that it’s a difficult problem because we want to cover everything, but in some cases we may need to lose some flexibility in order to guarantee that this more standardized.”