Andy and Kevin discuss how Genie evolved from Advance within the AIQ platform, focusing on a scaling problem called "offshore delusion" where testing costs rise faster than volume. Newer AI apps can cut labor costs, speed time-to-market, and improve accuracy.
It sounds to me though, one of the things you're getting at is the sheer scale of the problem means that you end up with this notion of what I call the offshore delusion. A unit of 1 is cheap, a unit of 10 less cheap. A unit of 100, not cheap. A unit of 1000 definitely not cheap. But the real element is, yeah, we can reduce the labor cost element, but we can get applications to market with greater accuracy, faster, we can reduce business risk and we can also start developing new, better applications even, even quicker. So there's a, there's, it's sort of a multidimensional benefit that you're bringing to the market. How did that grow out of Advance? Because I'm really interested in this idea that you had advance, you're doing well, you're really having an impact on the industry. And then suddenly we tend to look at these new AI apps like Genie and go, they're going to eat the lunch of the, of the software QA community, really. They're going to make their jobs more efficient, better leaner and elevate them to some. But it also results in really significant change for a business like Advance.
How does that work?
How do you go from advance to Genie?
Well, it's still the company atfance and Genie is a new offering within the AIQ platform. Right. But Genie is what customers have been asking for. So if you talk to CIOs, they know where they're spending money and they're spending money on regression testing of existing applications in the thousands of applications. And each of those has a thousand or more regression tests that, you know, end to end tests that have to be done, mostly manual, some of them are automated. And so again I think you look at how do we obsolete one of our own products and one of our products is Web Designer as well as Mobile Designer and Desktop Designer. These are really state of the art recorders, fastest, most capable, most self healing in the industry. We have to be. Actually it was Vinod Khosla. Not everyone watching this will know who that is. But I think it was Vinod who said if you're going to do something in AI, it needs to be at least one or two orders of magnitude better, faster, cheaper than what you're replacing. Otherwise why would people take any risks? They just keep doing what they're doing. And that's part of the problem with the recorder, right? The recorder is five, maybe ten times better faster than Selenium. And it didn't quite, it didn't replace Selenium. I mean Selenium still used more today than it was then, so we had to be 10 to 100 times better. Really two orders of magnitude better than the recorder. Right. And we make the world's best recorder, in my opinion. We could have great industry arguments about that, but that's what we had to do. So that's what we set out to do. And we did it in conjunction with customers. What is your biggest pain point? And you will hear a lot of people say, well, I wish AI could write my test cases from my business rules. You know, we already to a great extent do that with AI script generation. That's already a feature in advanced since 28. We create brand new test cases and brand new scripts for which they didn't exist before with your validations and with your BDD validations. Now Genie says you have all of these regression tests written as test cases. All of them. Thousands, millions in some cases. If you could push a button and let AI automate those, of course, not only would you do it, you'd say goodbye to your manual testers and 30 or 40 manual testers is run thing. But how about when companies ultimately deploy themselves and through their sis, Right? Thousands and thousands and thousands of them. As you said, one not too expensive, 12,000. Very expensive. Right. And that's the interesting thing. So that's what we're looking at doing. We are obsoleting the recorder. We are obsoleting selenium because it's 100. This is 100 times less expensive than anything you can do with either recorder or Selenium or manual testing. It's two orders of magnitude better than any of those in terms of cost. If I'm a CIO and I've got digital transformation and I'm bringing AI in, the number one thing CIOs have said on the record, including to me, but on the record in the press, is I need to see the cost savings. If I do not see cost savings, I'm not going to pay for it. So you've seen all this little bit of backlash with Copilot. Microsoft Copilot is great. We're all using it. We're all spending $20 a month. And after a year of it, CIOs are canceling. Why? I didn't lose ahead. I didn't save any money anywhere. Maybe my people were a little more productive. I couldn't measure it. If I can't find the savings, where's the money coming from?
