Comcast Delivers aIQ Application in Just Eight Weeks, Driving Higher Customer and Employee Satisfaction Scores

The Comcast brand is synonymous with home entertainment and telecommunications. While Comcast is well known for its products, it has evolved into a technology company focused on leading new development initiatives to enhance its products, solutions, and customer experiences. Technology is used as a business enabler across the organization.

For Comcast’s leadership team, improving and optimizing customer service has become one of the focal points of its technology development strategy. “As technology trends develop and new products become available, we’re able to broaden the range of what we can do for our customers, which drives new ideas at the business level,” says Jason Michener, director of software development and engineering at Comcast. Over the past few years, optimizing customer service products and offerings through new technology initiatives became a top priority for Comcast leadership and its engineering teams.

“We want to upend the traditional model for customer service and shift the conversation for customers. We don’t want to wait for our customers to reach out to us. We want to use technology like artificial intelligence to move from a reactive to a proactive model of engagement,” says Michener. “Given the digital age we live in, customers are rightly expecting so much more from their products and services. We want to approach customer service as a product and constantly work to improve our customer service offerings to meet and exceed evolving customer expectations.”

As the company's business and engineering teams began to think about measuring product output and impact differently, the development culture at Comcast started to evolve.

“The first question for our teams became, how does our approach to development impact the customer in a consistently measurable way?” says Michener. This cultural shift led to a marked investment in research and development of Comcast’s aIQ program, which focuses on using deep learning and emerging artificial intelligence (AI) and machine learning (ML) technologies throughout the company. Comcast has endeavored to use these technologies to transform their approach to customer service.

Consumer research demonstrated that there were enormous opportunities to improve customer experiences through a digital, data-driven, and predictive analytics-focused approach.

Compared to traditional means of customer service and communications—such as calling a 1-800 number or visiting a Comcast location—AI and ML technology could create easier and faster customer service experiences. The team found that they already had many of the critical components—such as Comcast’s engineering and UI resources, customer interaction channels, and internal leadership support—necessary to build a platform that could take advantage of deep learning to drive virtual, self-service assistance to customers across nearly all of its digital channels.

The aIQ platform emerged from this research. Taking advantage of natural language processing and using AI, the team felt the aIQ platform could fundamentally change the customer experience while also changing the way internal teams interacted and collaborated.

With the aIQ platform, Comcast sought to drive proactive monitoring and predictions through ML. Using AI, the aIQ platform provides interactive assistants across all products and channels. This increases agent empowerment through the synthesizing of rich history and event data to provide new insights and solutions and enhances the ability to learn new skills and apply rich context when evolving diagnostic data through an extensible framework.

Given a rapidly approaching end date with another customer service contract provider, it was business-critical that Comcast build and release aIQ within a specific timeframe. With a strict deadline of 8 to 10 weeks set to build aIQ, Comcast got started on development. Comcast chose to engage Pivotal to develop and deploy the aIQ application on Pivotal Platform running on AWS. Comcast had already engaged Accenture and Pivotal to migrate large enterprise services onto Pivotal Platform. By having these monolithic backend services already broken down into microservices running on Pivotal Platform, developers at Comcast were able to build the necessary backend service integrations for aIQ more quickly.

"Other teams at Comcast had used Pivotal, and we had engineers already familiar with using Pivotal Platform, which was enormously helpful,” says Michener. “Pivotal provided us with guidance in terms of industry and market trends and were excellent partners for us to work with as we developed our strategy to build, test, deploy, and scale the aIQ app quickly to meet business objectives.” Throughout the eight-week development timeframe, Michener and the team worked closely with Pivotal’s team. Pivotal’s thought leadership and platform architecture advising capacities ensured that the Comcast team built a superior cloud architecture for the volume and scale aIQ needed to support.

“We closely collaborated with Comcast’s team to figure out exactly what was needed and what we considered critical to the development and ongoing management of the aIQ application,” says Chris McCune, global account director at Pivotal. “Together, we embraced agile development methodologies and best practices to accelerate delivery and improve software quality.”

The aIQ app runs on Pivotal Platform on top of AWS and is powered by many AWS-native services. "The capabilities AWS provides are fantastic," says Michener. aIQ's architecture is deployed in 4 AWS Availability Zones (AZs) across 2 AWS Regions, providing high availability and disaster recovery mechanisms.

aIQ’s Amazon Elastic Compute Cloud (Amazon EC2) architecture consists of 636 Instances and 1,400 Amazon Elastic Block Store (EBS) volumes running on 2,500 central processing units (CPUs) and using 50 transaction processing systems (TPS), scalable to 500 TPS in minutes. Comcast stores over 2 petabytes (PB) of data on Amazon Simple Storage Service (Amazon S3). The application also takes advantage of AWS Trusted Advisor, Amazon Route 53, Amazon Aurora and PostgreSQL, and AWS Lambda for low-cost operational tasks. aIQ uses a dedicated AWS Direct Connect link to communicate with Comcast enterprise services that run on Pivotal Platform in Comcast data centers.

Pivotal Platform provides Comcast with built-in security and automates time- consuming operational tasks such as patching and upgrades and provides easy log integration for faster troubleshooting. For Comcast’s developers, Pivotal provides automated pipeline capabilities and a single command to deploy code with no downtime.

With over two years in production, aIQ on Pivotal Platform and AWS has reduced Comcast’s operational costs —with only three DevOps engineers out of a team of 45 responsible for platform and application management. The application has experienced minimal operational impacts, and Comcast has reduced its time-to-market from months to weeks.

“Since rolling out the Xfinity Assistant powered by the aIQ platform, we’ve found a 40 percent reduction in customers calling us after using the app. Our Net Promoter Score, which is a key tool for measuring customer experience, has improved and continues to improve,” says Michener. Comcast’s Net Promoter Score (NPS) has, on average, increased 20 points since 2016. “Customers are happier working within our digital channels,” says Michener. Comcast recently won a CIO 100 award in technology innovation for their Xfinity Assistant, powered by the aIQ platform.

For Michener, the impact of using Pivotal Platform and AWS for aIQ begins with how the platform and infrastructure impact internal development speed, freedom, and collaboration.

“Utilizing products and technology like AWS and Pivotal Platform and expertise from these organizations gives our people more time to collaborate, innovate, work together, learn from each other, and embrace a DevOps lifestyle,” says Michener. “There is so much automation and reliability built into these platforms that we leverage so my team can spend more time thinking about and collaborating on new development ideas and doing things that provide real, measurable customer value.”

Teams focused on cloud development have seen a 22 percent uplift in productivity versus non-cloud development teams. Furthermore, these teams report a 15 percent increase in employee motivation scores. Using Pivotal Platform and AWS for aIQ not only helps Comcast drive better customer outcomes but has also transformed the customer engagement organization into a high performing, dynamic, and motivated software development organization.

“What excites me most about these platforms is what our people can do to lead in the technology development space," says Michener. "Not only are our engineers able to take better advantage of emerging technologies, which helps them advance their expertise and their careers, but they are also happier at work. To see them excited to come in every day and have the freedom to innovate, talk, and collaborate, to love what they do and leverage technology to make real impacts for customers—that is what excites me about coming in every day."