The Center for Data Innovation spoke with Sean Barry, chief executive officer (CEO) of Kobiton, a platform that uses AI to automate and fast-track testing and launching new mobile applications. Barry discussed Kobiton’s AI-augmented testing, how AI helps overcome the main challenge for companies testing mobile apps, and how advances in generative AI will shape the company’s future.
Martin Makaryan: What is Kobiton?
Sean Barry: Kobiton is a platform that supports manual and automated testing for mobile applications. Developers, testers, and user interface engineers all use our platform to ensure that the apps they are building perform according to their desired specifications for the end users. This means anything from the visual appearance of an app to how an app might display some buttons when you use an iOS device as opposed to an Android one. What sets us apart is that we are using AI to offer “augmented” testing, which allows engineers to perform more tests in a shorter time by, for example, automating test script writing in a programming language that engineers currently use.
We have customers on every continent—except Antarctica! They are typically global enterprises where mobile apps are critical for customer satisfaction. This list includes large banks, airlines, government agencies, and companies where mobile is their product, like rideshare apps.
We offer three deployment options for our customers. For highly regulated institutions, like government agencies, we deploy our software and hardware on government-owned devices on their premises for maximum security. We also offer on-premises deployment for non-government clients, which uses our software with customer-owned hardware. Lastly, we offer a fully cloud-based option, where we manage the testing process in our large data center in Atlanta, Georgia. If customers choose this option, they can test their applications on thousands of smartphones that we have available for testing.
Makaryan: What is the process for a customer who wants to run their app tests through Kobiton?
Barry: Once a customer joins Kobiton, they install their apps on real devices—either at their location or in our cloud environment. We are big believers in testing on real devices because of the variance in how apps perform on different versions of smartphones and operating systems.
Afterwards, the Kobiton software leverages AI to help enterprises identify issues with their apps. Customers define and write their own tests, and our platform facilitates the process they have outlined for testing. This includes visual, performance, and regression tests. Regression testing refers to tests that verify that previously tested apps will run as expected after a change.
Makaryan: What makes your platform unique?
Barry: I would highlight three areas where our platform excels and sets us apart. These three differentiators are speed, flexibility, and AI integration. The Kobiton platform is the fastest in the industry for both manual and automated testing. Speed is crucial for many companies who need to constantly release new apps or features in existing apps. In terms of flexibility, our users can easily share test information across their entire software development lifecycle team, allowing engineers to easily move data between their teams.
Lastly, we are proud of how we have leveraged AI models to simplify the entire testing process, including by enabling less technically proficient users, like staff who can’t code, to perform complex testing tasks. AI allows people to do more thorough testing with the time they have. Time constraints are a constant challenge that we see our customers face when they want to test a new app or a new update. Often, they do not have enough time to thoroughly test before a release, which can lead to serious quality issues. As a result, a company would then have to focus on fixing those quality issues rather than adding new features to provide a better user experience. AI helps us overcome the time constraints to a great extent. For example, Nike decreased the time it took them to test an app from 160 hours to just 12 hours using the Kobiton software.
Makaryan: Can you discuss specific ways in which you leverage AI to enhance the platform?
Barry: We use AI in many ways, but I would highlight two specific areas where AI is a game changer for us: visual testing and script generation. First, AI algorithms perform ocular reading of what is happening on a user screen and translate it back to the user, checking for things like fonts, text, colors, and other features that are important for a smooth app experience. Second, we have an AI model that generates scripts in different programming languages that engineers use to run tests. This feature enables efficient testing across multiple device types and operating systems.
Makaryan: What is your vision for Kobiton’s future?
Barry: Our growth opportunities push us to constantly brainstorm new ideas and use new technologies to help customers test their apps thoroughly and in a manner that suits their needs. As more services move to mobile platforms, including retail and field services, we see significant growth potential. Our goal is to help customers deliver quality apps in a short amount of time, regardless of where they are in their journey, both in the United States and globally. This means that future advancements in new technologies, especially in generative AI, will be important for our growth. We are looking forward to more innovation that will allow customers to test better and even faster than what we have already been able to accomplish with our software.