Satish Mallela is Vice President, APAC at Applitools and has over 20 years of experience helping organizations transform the way they build, test, and deliver software. His career spans software development, quality engineering, solution architecture, pre-sales, post-sales, customer success, and executive leadership, giving him a unique perspective across the entire software delivery lifecycle.
At Applitools, Satish has helped enterprises across APAC adopt AI-native quality engineering, enabling teams to embrace VisualAI, Autonomous Testing, and MCP-powered development with confidence. He is passionate about turning complex engineering challenges into practical, scalable solutions that improve software quality, accelerate delivery, and maximize the value of AI.
The Verification Gap: When AI Says It’s Done, Can We Trust It?
AI agents now write code, generate tests, fix failures, and ship features with little human review. We are getting very good at making agents build software. But are we as good at knowing that what they built is right?
An agent can pass every test it wrote for itself and still miss what no test thought to check: a button that disappears on one browser, or a layout that does not match the approved design. The code compiled. The tests passed. The agent said, “done”.
This is the verification gap. For testers and quality engineers, it means more to check, in less time, with the same team. Through short recorded demos, I will show where the gap hides and how an agent can get independent evidence that its work is right, using VisualAI and the Applitools MCP.
Takeaway: give your agents an independent verification layer, so “done” comes with evidence.
