Category: Advanced Analytics and Data Science
At Quicken Loans, we believe that optimal model function is best achieved by supplementing our models and data scientists with an advanced suite of supporting technology. We call this our “model ecosystem” approach, which quickly grew beyond a simple virtual assistant for offloading reporting and other mundane tasks. It has become a game changer that revolutionizes our approach to model building. Today, it is our go-to that orchestrates ensembling, model evaluation, and model selection. It suggests which models to retrain, which to roll back, and what portions of the model ensemble to focus on. It also forecasts model accuracy and continuously updates estimates of its own accuracy and the likelihood of further deviation. Our clients can obtain visual explanations of model performance down to the attribute level.
The benefits of developing our model ecosystem approach are manifold. We can take on much larger workloads with ease. During a recent market event, the system provided early alerting that saved the company millions of dollars, conservatively estimated, over a 30-day period. Similar examples abound over the 27 months the system has been in production. The path is clear for developing AI capabilities that will enable the system to learn the rules of great performance and become the ultimate complement to the human data scientists on the team.
Category: BI and Analytics on a Limited Budget
Club Assist LLC.
Club Assist North America is the preferred supplier for the American Automobile Association (AAA) and the Canadian Automobile Association (CAA) Mobile Battery Service Program. Club Assist's Smart Data platform synthesizes and analyzes multiple rich data sources to deliver actionable and reliable business information internally and externally. Dashboards and reports can be tailored with high-level performance metrics for executives, or with intensely granular data for hands-on field management. The platform benchmarks performance across clients, track initiatives, manage resources, and assists customers in driving their businesses. External Club customers can extract reliable, robust data. Club customers then reference this data to hold providers and team members accountable for testing and sales conversion.
Smart Data uses Kimball life cycle methodology and snowflake schema to allow Club Assist to efficiently integrate diverse data sets and generate complex analytics. Leading the automotive industry, Club Assist's use of Smart Data aligns battery test data with dispatch and emergency roadside call information, offering keen insight regarding the comprehensive roadside process. This insight produces greater accountability at all business levels, driving impressive revenue gains since inception. Smart Data processes 12 million data records per year, driving better results for 23 Club customers and servicing 27 million members.
Category: BI, Visual Analytics, and Data Discovery
Vision 2030 of Saudi Arabia endorses the digitalization of systems and processes, and has been adopted by Saudi Telecom Company. STC’s DARE Strategy (Digitize, Accelerate, Reinvent, Expand) has re-invented STC’s customer experience to a world class level through personalization of every interaction, as is reflected in BI Analytics & Technologies road map. This means we have championed our traditional BI implementations and services, and kept our product fresh by heeding current and upcoming trends, including integrating location intelligence, contextualized marketing, microsegmentation, loyalty analytics, big data analytics, social network analytics, and machine learning with our EDW platform. Through this, our business users have achieved greater revenue, employee retention, cost optimization, and company growth.
STC has a state of the art business intelligence architecture with an exceptionally responsive predictive analytics and data discovery platform. Its visual analytics is the core competency of the organization, giving it the capability to serve a diversified range of business users with pragmatic solutions to their analysis needs. Our focus is to deliver interactive and insightful dashboards and visualizations, and to provide actionable insights to the business.
Several initiatives powered atop our BI platform—including smart investment, location-based campaigns, sentiments analysis, after-sale services, self-service, mobile analytics, and loyalty analytics—have proven fruitful for STC and established us as a market leader in BI and data visualization services. These services are shared with partners to achieve their respective goals and KPIs in a more efficient and innovative manner.
Category: Data Management Strategies
Solution Sponsor: Talend
Anheuser-Busch InBev (AB InBev) is the world’s largest brewing company and one of the largest and fastest-moving consumer packaged goods companies. Today, AB InBev is a $55B worldwide business, made up of a portfolio of nearly 500 global and regional beer brands across 100 countries, including notable brands such as Budweiser, Corona, Stella Artois, Michelob, and Leffe, as well as leading craft brewers. The company has transformed through a series of mergers and acquisitions into a sprawling estate of technology systems and vast amounts of data.
Almost two years ago, Harinder Singh was appointed to lead data strategy, architecture, and governance with the mandate to build a global data culture and guide AB InBev along its data journey. Singh, an industry veteran with 15 years of experience leading data and analytics for Fortune 100 companies, determined to integrate all the company’s data into a unified view and enhance its analytics capabilities. This unified view is becoming the backbone for AI and blockchain initiatives.
AB InBev uses AI for use cases that range from optimizing promotions to image recognition for planograms. The company uses AI in combination with POS system data to gain insights into consumer preference, and understanding assortments, and how to combine beverages to optimize them. In supply chain and logistics, the company is using AI to optimize routing for fuel and traffic, as well as the best time for the customer to receive the order. AB InBev also uses data and algorithms for ethics, compliance, and fraud detection.
As part of the data management modernization, Singh chose to build the data architecture on the cloud for scalability, replaced or consolidated all data integration tools into one for dynamic and rapid development, and prioritized reusability over customization. AB InBev partnered with Microsoft Azure for cloud services and Talend for data integration. Within two years, Singh built the data team from the ground up, distributing it across three continents to support the global foot print of the business.
Category: Emerging Technologies and Methods
Solution Sponsor: BMC Software
BMC had a well-established enterprise data warehouse and reliable metadata layer with optimized, predictable, and well-understood data. However, it was no longer enough to have data and content delivered by IT; business users wanted to be the owners and creators of their analytics content, driven by changing skill sets and quickly evolving requirements.
IT shifted from traditional gatekeeper to enabler. We built an ecosystem that accommodated more advanced use cases and frameworks that allow business users to blend trusted enterprise data with their own emerging sources and departmental data. Combined, they form our data democratization strategy.
BMC now has an innovative, collaborative environment that meets a variety of self-service requirements. With our secure architecture, certified enterprise data can be accessed through a user-friendly metadata abstraction layer, while advanced users can still access raw data. Data can be enhanced, transformed, and visualized to meet the needs of each department. Data prep tools can cleanse, enhance, and transform the data; users can experiment with data and solve new business problems faster and with greater agility.
This is only the beginning. We expect data democratization to continue to evolve within BMC and have already seen organizations effectively put the ecosystem to new uses by creating machine learning and predictive analytics projects.