Pramin Pradeep on: AI-Driven Quality Assurance | Ep 1206

0:00Welcome to Corazon Technologies, home of the Digital Executive Podcast. Do you work in emerging tech, working on something innovative, maybe an entrepreneur, apply to be a guest at www.corazon.com forward slash brand. Welcome to the Digital Executive. Today's guest is Promin Pradeep. Promin Pradeep is the co-founder and CEO of BotGage AI, a US-based, autonomous QA as a solution company redefining how modern software teams ensure quality at engineering speed. Using QA as a solution model, BotGage redefines quality assurance for fast growing engineering teams. It combines AI-native testing agents with forward deployed QA pods to continually create run and maintain and end-test with owned quality outcomes. With over a decade of deep experience in low-code ecosystems and enterprise QA transformation,

1:04Promin has built his career at the intersection of automation and scalable software infrastructure. He previously helped scale a high growth startup from inception to 3 million in revenue, contributing to its acquisition by SOS labs. Promin has worked with leading enterprises including Adobe, Infosys, and Uncork to streamline software testing and quality operations. Look at afternoon, Promin. Welcome to the show. Yeah. Thanks, Sam. Thanks for inviting me. Absolutely, my friend. I appreciate it. And I know you're generally based out of the Bay Area San Francisco, California. I'm in Kansas City. Today you're traveling. I understand. So we'll just jump into it here. Promin, let me ask you first question here. You described BotGage as an autonomous QA as a solution company. Whoops, fundamentally broken in traditional quality assurance, models that require rethinking quality as a service rather than a function. Yeah. Great.

2:05So to understand that, we have to understand the history of the Coltee and Serenade. Coltee actually has speeds, right? So here, if you take before 10 years, every application which came into production, the release cycle, that is, every changes was happening once in six months, once in three months. However, as the development progress, the customer expectations started increasing. They want more and more functionality. They wanted to get into the latest update, the use of everything further coming into picture because of that, the competition increase and the release pressure started increasing for all the SaaS companies. So from once in six months, it started releasing into once in two weeks, then once in a week, and once in an hour. So because of that, automation become more prominent. Before 10 years, if you ask anyone, they'll try to say that, okay, well, if you company start doing automation, it may be true, they write script, I mean, it was sending in my open source. Our right now that need of our is for the QA to cope with the start of release cycles

3:13with prominent automation needs. The Bleacher Report app is your destination for sports. Right now, the MBA is heating up, March Manus is here, and MLB is almost back. Every day there's a new headline, a new highlight, a new moment you've got to see for yourself. That's why I stay locked in with the Bleacher Report app. For me, it's about staying connected to my sports. I could follow the teams I care about, get real-time scores, breaking news, and highlights all in one place. Follow the Bleacher Report app today, so you never miss a moment. For that, only the AI can be built into it, and it should input most regress way into any SAS Echoes of Confondante. Thank you. I appreciate that. And you're absolutely right. I was a developer early in the days, and looking back, QA was pretty strict, right? We had some really structured release cycles, and as you talked about, release cycles were not very frequent. But as the demand for more enhancements, more functionality came about, QA, it was hard

4:19to keep up with the QA, and I totally get that. But what you've done is really brought an autonomous level of quality at a faster scale, a faster turnaround. So I appreciate that. And Promen, BotGage combines AI native testing agents with four deployed QA pods. How does that hybrid model improve speed, reliability, and ownership compared to purely automated or purely manual testing approaches? Yeah. So I have been in this space for more than 10 years now. So seeing the journey from an open source, automation, to local automation using NLP through the AI code, right? Right now, what is happening? Just finding over the agents or AI agents through the company does not want to look, because every SaaS or every software is different and go through a lot of customization. So the learning process will be very important for any agents which will integrate into their ecosystem.

5:20So for that, what we have done different is like, we enhance our agents and deploy into their ecosystem, however, a forward deployed engineer, as to monitor any monitor and analysis of the agents, such a way that whether it is learning in the right format, whether it's processing the boundary condition or not, all those things has to be constrained. All those constraints have to be kept in mind. That's why we not only deployed agents, there's a forward deployed engineer, which monitors end-quent operation of the agents and also provide inputs, even deviating from the part. So that's why it is very important to our book agents, the human and the look to make sure the customer is able to get the right output. So here the output is nothing but an increased coverage in the short period of time. Just giving you numbers and plays, right? Consider they just move with the open source core, local tool, which is available in the

6:20market. Up please, the customer is going to take an on four to five months to reach 80% of coverage. With the agents which are built and the human and the look will be able to do it in two weeks of time. So that's the kind of onboarding of test cases and the coverage will be able to implement in any ecosystem compared to the traditional methodology, which is there in place. So it's an increased efficiency, increased coverage, which will support their release cycles and they can reduce their release cycle from two weeks to two days with the kind of deployment which we do in that infrastructure. Thank you. And you're right. You highlighted that hybrid model, efficiency, quality, faster terminal and times, that learning process for AI agents does take some time to learn and you want to make sure that it's accurate. Like how you use this hybrid model of having that human and a loop, you mentioned that forward deployed engineer to actually monitor and make sure that it stays within its parameters.

7:21So I appreciate the insights. And prominent, many engineering teams prioritize shipping features quickly. Why do you believe the next decade of innovation will be defined by autonomous quality infrastructure rather than faster coding alone? Yeah. So it's a very interesting question, Brian, because Brian, if you see him from a developing standpoint of you, there are multiple companies addressing the problem statement, like some call it invite coding, right, increase the level of coding enhancement to prompt, everything is just happening. However, once the speed of coding is entertained by the ecosystem, right? However, they're not able to release in the production because of the way they're doing the bottleneck. Because customers, you as you already know, they don't accept bugs or any slow broken. Once that is done, they'll just shift to competition. But that becomes the most important point. And end of the day, it's not about writing the code.

8:21They release into the trial and customer should start using it. For that end to end, the patient has to be done. That's why it's important to have an autonomous QA framework integrated into any infrastructure to support these release cycles in place. So what I want to tell the thing indeed, yes, writing the code is much important. However, the most important part is shipping to the customer to get the feedback loop established. For that, you have to tightly integrate the autonomous QA also into the framework. Thank you. Appreciate that. You did highlight some things that are happening at all different levels of sizes of organizations. But the speed of AI agents with this low code, this no code, QA is definitely the bottleneck right now. People can't wait for things. It's just kind of how we are in our human nature. But you highlighted the fact that having that develop and integrate that autonomous, those autonomous agents in that QA process will help speed this along.

9:25And obviously we want to have quality output, of course. So I appreciate that. And the last question of the day, as by gauge scales following its recent funding round, what will separate autonomous QA platforms that truly deliver outcomes from those that simply layer AI on top of legacy workflows? Yeah, so it's like you cannot modify the existing infrastructure and trying to make it AI by just adding a layer to it, right? You need to build everything from scratch, especially if you're going for AI first companies. For example, the traditional players where their code is already written and they're not able to even let me just want one of the major pain point in automation that is self-feeling maintenance. When an element moves around or changes or the flow changes, right now the traditional framework cannot cope with it. This kind of autonomous being the nature, that means the initial algorithm has to be written

10:29for AI. And that's why it's very important for any AI for companies to build from scratch and adding in there on top of the existing insura. That's on board gauge is the board gauge is an AI bond company where we just started in the area, the all the code, not only the element infrastructure layer, but also the algorithm to support that has been written from scratch to enhance the agent from the first learning approach, how can it refine from our outstanding refine to the most extreme level of learning the end-to-end application in a short period of time. Thank you. Really appreciate that and you did talk a little bit about, especially with your company, but AI companies in general and AI platforms, it is best to build everything from scratch if you're building that type of infrastructure. As you know, there are problems with adding multiple layers or adding AI layer just on top

11:30of legacy workflows, obviously, that is going to add more complexity and more problems down the road. So again, I appreciate you teasing that apart for us. And Promen, it was such a pleasure having you on today and I look forward to speaking with you real soon. Yeah, thanks for inviting me to this podcast, right? And I appreciate it. Yeah. Bye for now.