September 28th, 2026 by Kendra Stansel
Between the rapid rise of AI agents and the pressure for faster release cycles, modern testing teams are standing at a major turning point. At STARWEST 2026 in Anaheim, Inflectra took the floor as a Gold Sponsor to help enterprise software leaders navigate this transition safely. Our goal for the week was simple: help QA organizations move past the AI hype and adopt real-world solutions that keep systems trustworthy.
Here are the biggest highlights from our time on the Expo floor and in the session halls.
Inside the Booth: Tackling Script Rot & Manual Overhead
Throughout the event, our booth served as an interactive hub for engineering teams eager to eliminate script maintenance headaches and tackle testing overhead. Attendees gathered at our SureWire.ai station for an exclusive look at the early-access beta, where they got hands-on experience identifying edge cases and closing the coverage gaps that traditional automation fails to catch. At the same time, we demonstrated how Inflectra.ai accelerates the entire test management lifecycle by automatically building risk profiles, generating automated test scenarios, and deploying self-healing scripts that adapt to UI changes in real time.
Beyond the tech, STARWEST gave us a fantastic chance to build deeper relationships across our ecosystem. Being located just a few booths down from our long-time partners at Checkpoint Technologies made it easy to collaborate all week. We capped off the conference by hosting a joint dinner with the Checkpoint team, giving us a dedicated space to connect over great food and honest conversations about where QA is headed. Those off-site moments served as a great reminder that while advanced tools are crucial, strong partnerships and shared expertise are what truly drive successful transformations.
On Stage: Navigating the Realities of Non-Deterministic Software
A major highlight of the conference came during CEO Adam Sandman’s presentation, Engineering Trust in a Non-Deterministic World. Adam broke down the crucial distinction between preparing for a future of fully autonomous AI and upgrading the tools QA teams rely on right now. While building frameworks for tomorrow's intelligent software is the long-term goal, he emphasized that test leaders immediately need practical, AI-assisted tooling to make today's test cycles faster and more resilient.
To prove it, Adam walked the audience through a series of live, hands-on demonstrations showcasing Inflectra’s current AI ecosystem in action:
-
Rapise AI Translator: Adam showed how Rapise uses smart translation capabilities to streamline script creation and eliminate maintenance headaches by automatically adapting to application changes behind the scenes.
-
Spira AI: Attendees saw how embedded conversational AI inside Spira simplifies test management, enabling teams to query project data, uncover insights, and instantly generate Gherkin scenarios.
-
SureWire in the Wild: In one of the most compelling moments of the session, Adam put SureWire to the test against a complex travel airline chatbot, demonstrating how Inflectra tools can successfully validate non-deterministic, conversational interfaces that routinely break conventional test frameworks.
By grounding his talk in live workflows rather than abstract concepts, Adam gave QA teams a practical roadmap. You do not have to wait for the future to build resilient, AI-empowered testing processes because the tools to protect and accelerate your releases are available today.
5 Things We Learned at STARWEST 2026
Outside of our own presentations and booth conversations, the conference was packed with valuable perspectives from top quality engineering leaders. Here are five major themes and takeaways that stood out to our team:
-
Evaluation and governance have replaced simple "prompt engineering." The conversation across technical sessions shifted dramatically from basic AI prompting to rigorous evaluation frameworks. With talks focusing on measuring faithfulness in RAG systems, managing non-deterministic agentic workflows, and establishing metrics leaders can trust, the consensus was clear. Testing AI requires structured evaluation pipelines, strict governance, and traceable evidence to prove that model output remains bounded and reliable.
-
The shift from static automation to intent-based, agentic delivery is here. Multiple tracks highlighted the industry's transition away from traditional, fragile script maintenance toward intent-based testing. Sessions explored how autonomous agents and "QA swarms" can handle end-to-end execution across complex toolchains. Rather than spending sprints updating static locators, testing teams are stepping into the role of orchestrators who define high-level intent and let AI agents manage the underlying execution paths.
-
Quality Engineering is taking over as the architecture layer for AI execution. As AI writes more of the core application code, QA is no longer just a gatekeeper at the end of the sprint. Industry leaders emphasized that quality professionals must lead the way in specifying, validating, and establishing the foundational rules for AI execution. By bridging the gap between technical program management and deep testing logic, testers are becoming the primary stewards of system reliability.
-
Validating non-deterministic interfaces requires an entirely new playbook. From testing travel chatbots to evaluating dynamic LLM applications, standard pass or fail assertions are no longer enough. Speakers across several tracks shared practical techniques for probing non-deterministic systems for bias, hallucinations, and conversational edge cases. Securing user trust means expanding our toolsets to test systems that continuously learn, evolve, and respond in unpredictable ways.
-
Business value and ROI are the ultimate metrics for QA success. A recurring theme throughout the week was connecting technical quality directly to business impact. Whether through cost-cutting initiatives, SLO-driven test strategies, or aligning quality engineering with executive targets, successful QA teams are changing how they demonstrate value. Test leaders are focusing less on raw test coverage numbers and more on translating stability, speed, and risk reduction into executive confidence.
The insights and techniques shared at STARWEST 2026 are already transforming how quality teams tackle enterprise automation, and we want you to experience these capabilities in your own environment.
Take the next step in modernizing your test lifecycle today with a free 30-day trial. You can explore our core platform completely free, putting the intelligence of Inflectra.ai and the resilient self-healing power of Rapise to work for your team with zero risk or upfront commitment.
