Editor’s note: TELUS, a Canadian communications and information technology company, has transformed their approach to data science with Google Cloud services. Here’s how they’ve broken down data silos, accelerated data engineering tasks, and democratized data access.
As a dynamic, world-leading communications and information technology company, TELUS is always at the forefront of innovation. We have made significant progress with our digitization journey over the last few years, modernizing our systems and networks to create new and improved growth opportunities. However, in that process, we have inevitably accumulated vast amounts of important data across various systems resulting in data access challenges for our teams.
Our vision was to unite siloed data, democratize it across our organization, and enable our data science team to effectively extract meaningful, high quality insights to help with important business decisions. Partnering with Google Cloud, we’ve approached our cloud transformation in a way that allows us to unlock the true potential of data to create valuable insights and deliver exceptional customer experiences.
From siloed data assets to a single source of truth
We began this transformation by cleaning up and starting the migration of our siloed data assets to Google Cloud, aggregating all our data points into a common data layer built with BigQuery, Dataflow, Cloud Composer, Cloud Bigtable, and Cloud Storage.
Data governance has been crucial through this to ensure we have a single reliable source of truth for our data. We created a TELUS Metadata Repository to document information about our data assets (provenance, business description, privacy and security classification) in order to improve our team efficiency and streamline productivity.
Democratized data unlocks new use cases
Our partnership with Google Cloud has also helped us to democratize data across the organization, allowing each business unit to share their data with each other and collaborate more effectively.
At TELUS, our data spans beyond just the telecom industry across multiple business units such as healthcare, security, and agriculture. By bringing all those different datasets together, we’re seeing new use cases that help us improve the lives of Canadians. As an example, our Data for Good program was instrumental in helping track the spread of the COVID-19 virus during the global pandemic. By providing governments, health authorities, and academic researchers a platform to access strongly de-identified and aggregated network mobility data free of charge, the program assisted in initiatives to flatten the curve of COVID-19, reduce its health and economic impacts, and contribute to studies that could prevent or mitigate future phases of COVID-19 or other pandemics.
Unifying the data and AI lifecycle
Our data science team has made tremendous strides with Google Cloud services to reduce machine learning (ML) model development and deployment time. We have been testing very sophisticated compute instances on Google Cloud, and Vertex AI, to accelerate our journey by unifying the data and AI lifecycles – from data exploration, aggregation and cleaning to model building, training, testing and finally deploying ML models in production. In addition to the acceleration of ML model development and experimentation, with Vertex AI our data scientists will also be able to implement Machine Learning Operations (MLOps) to efficiently build and manage ML projects throughout the development lifecycle.
Moving to the cloud has not only accelerated our model development, it has also allowed us to innovate faster. We transitioned from a waterfall to an agile mindset early on, but we needed an even faster framework to trial many ideas in just a few hours. We’re trying to empower the team to rapidly test their ideas, accelerate their iterations, and minimize the impact of their failures. This has enabled us to determine within just a few days—as opposed to months—if a project will be successful or not and therefore, minimize wasted time.
Privacy and security remain at the forefront
As we grow our data science practice and use these tools more widely throughout our organization, keeping our data secure remains a top priority. We’ve established the TELUS Trust Model to reflect our commitment to protecting our customers’ personal information. To build trust with our stakeholders, we always use this data with respect and make sure that security and privacy is built into every step of our projects. Using Google Cloud allows us to retain complete control over our data and ensure that any information we use for analysis is always de-identified, so it can’t be attributed to any single subscriber. While Google Cloud provides Data Loss Prevention (DLP) service, it does so in a way that doesn’t slow down our time to retrieve insights. In addition, we leverage Google Cloud locations in Montreal and Toronto to help support data sovereignty requirements and ensure that our customer information never leaves Canada.
Data champions shape the TELUS culture
Since we’ve transitioned to Google Could, TELUS has also undergone a significant cultural shift. We’re driving TELUS to become a next-generation, insights-driven organization that creates valuable analytics to maximize business outcomes and deliver superior experiences to our customers. Moving forward, we are excited to continue leveraging our insights, AI skills and technology to create meaningful human and social outcomes and help build stronger, healthier and more sustainable communities.
Learn more about TELUS’ Data for Good initiatives and overall data cloud use cases here.
Cloud BlogRead More