STARWEST 2023 - Concurrent Sessions

API First Platforms adopt rapid application development lifecycle and high velocity deployments. Quality assurance teams are stressed to reduce the automation backlog and have a limited window to catch up with speed of deployments. This session will discuss how Venkatesh's team at Oracle adopted a creative approach to accelerate test automation by adopting an innovative “Automate First” model. The core pillar of this test architecture is creation of acceptance driven component tests leveraging virtualized services by developers. This laid a solid foundation which assured validation of...

This talk will provide a glimpse into how continuous test pipelines co-exist and integrate with strategic low code test automation tools. Multiple testing teams, each owning distinct applications, are faced with the task of testing their own product as well as the interactions and behaviors with each other. Tight schedules and the challenge of finding the right skill set can be solved with the use of low code SAAS test automation tools. In this talk, Shai Srinivasan will share some of the challenges she and her team faced—how do these tools integrate with the enterprise's testing process...

Let’s be real, we are quite a ways out from AI taking over the world. AI doomsday apologists stay home! On the other hand, AI is changing everything about software creation. In this presentation, Kevin Pyles will take you on a walk down memory lane through waterfall, agile and DevOps practices, and will then introduce you to the future of software creation known as AI-Partnered Development. Let’s get practical with real-world examples of AI assisting designers, solving coding challenges and testing software right next to human testers. We will explore how AI is also involved today in...

Maintaining API test cases across multiple environments, for different load levels, and types of tests (eg. smoke, functional, regression) traditionally required separate tests be written and maintained. Test maintenance is expensive. If you have three lower environments that tests are run for, then you usually have at least three different tests you are trying to keep in sync. If you have a new API added for a microservice, then you have to modify and validate across all three tests at a minimum. If you are running different load levels for build tests verses load tests, for example, that...

JavaScript code requires effective testing strategies to avoid bugs in production. Code coverage metrics alone can be misleading and don't necessarily indicate that the code is actually working as intended. It can be easy to get high code coverage metrics, but not actually test anything meaningfully. To address this problem, this session will explore the concept of JavaScript mutation testing. Mutation testing introduces small changes or "mutations" to the code and then expects tests to fail. This technique helps identify areas of the code that may need improvement and strengthens...

AI can now generate an automation script as good as humans in seconds. It can teach you about an API or write documentation for your code. What is really going on at the forefront of building AI systems for test automation? Working at companies like test.ai and Google, Chris Navrides has explored the intricacies of building a robot to assist a manual QA team. Through large data systems, he has dealt with training machines to determine a bug versus a feature. Chris will show how to best leverage this new wave of AI to be competitive in the work force, and increase your efficiency. As well...

If you can't measure your software quality, you can't manage it. Cox Automotive implemented service virtualization for their Dealertrack product development to more effectively diagnose and address issues in the application. This addition to their test automation has improved the overall quality of their product lines by significantly reducing the escaped defects. Visibility into key metrics, like escaped defects, can provide insights into potential risk areas so teams can make smart choices about what tests to run to achieve the desired code and test coverage. Join this session as Roya...

How do you go about testing the immutability of deployed code? What control does one have over a decentralized test environment? How does one overcome the complexity of smart contracts and associated potential vulnerabilities? What can be done to ensure seamless integration and interoperability between the blockchain application and external components? As a tester of blockchain applications, Gomathi Ramalingam has encountered and faced many of these challenges, which are inherently due to the unique nature of blockchain technology. She believes that having a good understanding of these...

In this exhaustive testing world, it can be daunting to measure multiple metrics to achieve the end goal of higher product quality. Measuring metrics can quickly become a daunting task that can pose several challenges on the time and efficiency front. What are these two simple metrics that can get you to your finish line? How can teams use these metrics to bring about a process change to set high quality standards? In this session, attendees will take away simple, but effective, ideas that can help their teams to clearly define their product quality goals, and attach a quantifiable...

Software Development Engineers in Test know that page objects have long been used to make test automation more resilient. This talk evolves the page object pattern in ways that provide some really strong advantages over the traditional approach. Timothy Heck will demonstrate how to make page objects isolated from the browser automation technology, making it possible to switch from Selenium to Playwright (or the next big thing) without changing the page objects. This talk will also look at how to make self-narrating automation scripts where the simple act of interacting with a page object...

Integration testing in the staging environment is a pain. One bad commit from a dependency, and all your automated integration tests start failing. Consequently, developers rightly don't want to gate their code pushes on successful integration tests, and integration test results are quickly dismissed as unreliable, and ignored completely. Ephemeral environments change all of that. By using ephemeral environments as a key part of test strategies, not only has John Jenkins succeeded in moving staging dependent integration tests into the build pipeline, but he has also caught and prevented...

In many cases, it takes multiple metrics to get a complete picture of the status of a process, product, or service. A metrics dashboard accumulates a set of metrics associated with a decision. A dashboard allows the metric's customer to see, in a single collection, the relevant information. In this session, Linda Westfall will use the question, "are we done testing?" to illustrate an example of a metrics dashboard. She will also discuss each individual metric that makes up the dashboard and how they contribute information to help answer this question. These metrics will include testing...

#A11y, aka Accessibility is a hot topic nowadays. It is not only an important business priority for each digitally transformed company to increase user retention and sales revenue, but it also ensures that people with disabilities can use any digital product without any barriers. Many organizations started to focus on digital accessibility by investing tremendous amounts of budget in vendor or user testing post product deployment. However, quite a portion of this budget can be eliminated by educating their product teams, especially engineers to adopt accessibility automated test in their...

As software systems have become more distributed and complex, the "shift-left" movement brings reliability testing to earlier stages of development. Ensuring reliability goes beyond simple end-to-end tests. To ensure the highest levels of reliability, you must perform a suite of testing types. Incorporate contract tests to validate APIs; load tests for scaling predictability. But how to test for those inevitable failures? Let's learn from Chaos Engineering principles by incorporating disruptive behavior into your system before production. Join Paul as we learn ways to incorporate a...

New approaches in robotics are changing the way companies build and test software as it becomes increasingly device dependent. Companies like Apple, Google, and T-Mobile already rely on in-house robotics fleets to automate manual testing of mobile apps on devices. And that is just the beginning. Robots can perform all of the expected actions that a human would make on a device, including finger swiping, switching between various apps, toggling the notification center, and more. Using computer vision, a field of AI that teaches computers how to parse digital images and videos, robots can...

Many times accessibility activities occur late in a project’s release when it takes longer to figure out problems and costs more to fix them. Typically all you can do at this point is patch the problems rather than fix them properly. CGI Federal approaches accessibility with an agile mindset, integrating it throughout the iterative design, user story creation, development, and testing efforts. During this session, Pavani Gonuguntla and Stephen Omojuyigbe will focus on case studies and recommended best practices from where they worked with management and cross-functional teams to get...

The success enabler tools supporting business organizations within an enterprise are primarily driven by the underlying AI/ML powered solutions, data & analytics platforms, and enterprise applications suite. In modern data driven economy, the dependability (quality and timeliness) of data powering these business enablers, directly influences the success or failure of the business offering and adoption. The major factors for constrained data dependability can be attributed to limiting capabilities for validation of data availability, usability, reliability, relevance, and presentability...

The relentless pursuit of defects is critical to any industry, but how does that change when the software is being used on 300,000 lb steel machines in some of the most extreme environments on Earth? When the safety of both human life and enormous capital investments are on the line? What is it like to work with software developers who not only won't but actually can't use the end product as a user because the target is a rock drilling machine? We all know that "garbage in equals garbage out" and it is just as true when relying on sensor data to determine machine response. However, how do...

In an increasingly interconnected digital world, consumers of online services have become more savvy with ever-increasing expectations. The performance and scalability of these online services are critical to their success, and load testing is an important tool to evaluate these capabilities. Traditional centralized load testing needs to catch up in replicating real-world scenarios. Organizations should replicate actual user behavior by distributing simulated user traffic across multiple geographically diverse locations, enabling a more realistic evaluation of system performance and...