Agentic testing platforms promise to confirm your app does what you meant it to; but what happens when it gets there by a path you never meant at all? Agentic testing tools promise something genuinely new, agents that pursue a testing objective rather than following a fixed script, adapting as the application changes underneath them.

That's a real architectural shift, not just marketing. But if you tell an agent to "complete the checkout flow" or "finish the approval," how do you know it didn't get there by skipping a step, bypassing a validation, or taking a route a real user never would? Learn to ask the right questions before you trust an agent's definition of "success."

* What's architecturally different about agentic testing versus scripted automation and AI-assisted script repair.

* A practical framework ("the oracle problem") for evaluating whether any agentic testing tool can enforce what a test must *never* do, not just what it must accomplish ... the gap most evaluations miss entirely.

* Questions for any agentic testing POC or vendor conversation.

Rick Cavallaro is President of RJCadvisors & publisher of QAMeditations.com, a Massachusetts-based QA and test automation consulting firm. He brings 30+ years of enterprise test automation experience with deep domain expertise in Oracle EBS, Oracle Fusion, and other Complex line of business applications, including work on DoD programs. Earlier in his career, he served as Director of Technical Services Worldwide and Senior Product Manager at Empirix. Rick currently runs a fractional, multi-client consultancy and is actively evaluating and writing about the next generation of AI-native and agentic testing platforms.

George Nealon is Senior Vice President of Global Sales at Functionize, where he has led sales efforts since April 2021. Before Functionize, George served as VP of Global Sales/CRO at HiveIO and spent over three years at IBM as API Economy & Integration Manager, Worldwide. His background spans on-premise, cloud, PaaS, and hybrid-cloud solution sales, with particular strength in growing startups and building new markets for global enterprises.

James Abbott, Senior Solution Engineer at Functionize. James has over 8 years of experience in growing and leading software quality assurance teams. Led international teams in implementing and maintaining automated testing frameworks.

September’s 33rd Kick-off audience participated actively in thought-leader Robin Goldsmith's eye-opening “How to Make Remote Project Teams Jell.” Robin pointed out the shortcomings of the offered common definition that a team is a group with shared goals. Similarly, comments focused on mechanical challenges of remote-working, such as time zones, distractions, and shared understandings, but missed issues related to jelling. Robin identified these are necessary for but do not themselves jell a team. The essential but too-often missing ingredient for team jelling is leadership. Watch the recording under the Calendar on www.SQGNE.org.