In the age of large language models (LLMs) and Generative AI, data quality has become essential for telecom providers looking to harness AI’s full potential. Recently, leaders from SK Telecom and Rakuten Symphony shared insights into how telcos can achieve optimal data quality for AI applications, highlighting the importance of data governance, standardization, and bias mitigation in making AI-driven initiatives successful.
The session, moderated by Abe Nejad of the Network Media Group (NMG), featured Eric Davis – VP of AI Tech Collaboration, SK Telecom, and Sandeep Arora, Senior Vice President (APAC & MEA), Rakuten Symphony, who discussed data quality as a critical enabler for effective AI in telecom. Given that poor data quality severely impacts predictive accuracy, revenue, and even regulatory compliance, the leaders pointed out how issues such as incomplete data, biases, and inefficient data management can disrupt AI model performance, leading to inaccurate predictions and lost opportunities.
The importance of a solid data foundation for AI cannot be dismissed. The leaders highlighted that addressing data quality from the onset is essential to building reliable and scalable AI systems. A practical example is SK Telecom’s contact center, which has consistently ranked number one in customer service. High-quality data enables SK Telecom to maintain service standards, automate customer support, and personalize user interactions. Accurate, bias-free data allows SK Telecom to cater to various demographics effectively, ensuring that AI-powered systems reflect the needs of all users. Among the topics discussed was their governance program, “T.H.E. AI” (Telco for Humanity with Ethics), which promotes inclusivity, transparency, and reliability in AI applications. This program sets rigorous standards, ensuring that every AI project aligns with ethical practices and goes through a thorough review process before deployment.
Meanwhile Rakuten Symphony’s approach to AI-driven customer experience relies on its revolutionary cloud-native platform, consolidating data from multiple sources into a common data lake. This data lake supports applications like customer experience management, network automation, and customer data record (CDR) analysis, enabling Rakuten Symphony to deliver contextual, data-driven insights to enhance customer satisfaction and operational efficiency.
“Data quality is not just critical; it makes or breaks projects. If you have poor quality data that’s biased, unfiltered, or incomplete, you’re going to make poor predictions that could cost you customers and revenue.”
“Data is the bedrock of AI. If we don’t get the data foundation right, we’re at risk of building biased models and untrustworthy outcomes. Quality data is essential from day one.”
Both speakers emphasized the importance of governance frameworks in ensuring data quality and responsible AI use. In addition, standardization and validation are integral, including eliminating outdated legacy formats and preventing data quality degradation over time. The journey to quality data in telecom is as much about culture and mindset as it is about technology, urging telcos to embrace data quality as a shared responsibility across teams.
The session highlighted that while AI promises transformative capabilities for telecom, achieving quality data is foundational to its success. Telcos that prioritize data quality will be better positioned to innovate, improve customer experiences, and drive operational efficiencies in an increasingly data-centric landscape.
Watch the full interview here.
