Home to a handful of urban hubs and many rural communities spread across approximately 665,000 square miles, Alaska is vast to say the least. Bigger than the next three largest states – Texas, California, and Montana – combined, Alaska sprawls across a rugged area of ocean, rivers, mountains, glaciers, tundra, and even rainforests. Since 1979, GCI has invested more than $4.7 billion in its statewide network to bring Alaskans access to reliable broadband and wireless services, connecting communities to each other and to the world. Providing these services generates trillions of data elements from hundreds of thousands of sources. GCI and Databricks have teamed up to tackle this challenge head-on, leveraging the Databricks ecosystem to revolutionize how data is captured, analyzed, and understood. This partnership is not just about tools – it is about creating value for our customers and our business.
GCI began its partnership with Databricks six years ago, when it migrated its legacy data and Hadoop infrastructure to Databricks running on Azure. Since then, GCI has leveraged Unity Catalog to consolidate data access across platforms, improved data governance with Databricks Asset Bundles, and built ML models to reduce churn and guide network investment. GCI is now using newer features to revolutionize the way it manages data. One example is the use of Metric Views to streamline data aggregation and transformation. Another is the use of AI/BI Genie and Dynamic Dashboards to deliver information to stakeholders.
GCI uses Metric Views to calculate customer experience and network performance KPIs from huge telemetry datasets. They have delivered several orders of magnitude improvement in performance and an order of magnitude in cost savings, compared to previous methods. Unlike traditional aggregate tables, which can be slow and cumbersome, Metric Views enable rapid, efficient analysis—even when using complex data grouping methods. Furthermore, the decoupling of metric metadata from data engineering has also allowed for better division of labor. GCI’s subject matter experts can focus on defining measures, while engineers focus on pipeline management. This division accelerates innovation and enhances collaboration between teams. And because Metric Views are deployed with Infrastructure-as-Code, version control and compliance are easy to ensure.
Internal stakeholders such as Marketing and Finance want more than charts – they want to interact with data by posing questions in conversational language, just like they do every day with public LLMs. AI/BI Genie enables GCI to supply them with chatbots that limit information domains to those relevant to the topic of exploration. It also adheres to security policies set in Unity Catalog. AI/BI Genie allows GCI’s analysts to push out information more quickly and avoid the back and forth of creating successive visualizations.
Dynamic Dashboards allow GCI to build visualizations more quickly with assistance from AI-assisted coding. They speed up delivery of information to stakeholders by eliminating the need to pass data onto another visualization platform. And the Databricks One interface allows GCI to tailor the availability of Genies and dashboards according to its security policies.
Looking forward, GCI and Databricks will continue to partner to deliver optimal customer experience. Says Troy Goldie, Senior Vice President and Chief Technology Officer at GCI, “GCI seeks to continually improve our customers’ network experience, and information exploration plays a key role in that effort. Similarly, Databricks continually improves its platform and anticipates our needs. And they do so while helping us keep our costs down. GCI and Databricks look forward to ensuring that Alaskans have access to the same high-performance networks as New Yorkers or San Franciscans.”
Learn more about the Data Intelligence Platform for Telecommunications, or request a demo to see how your organization can drive competitive insights.
