MuukTest Emphasizes GenAI as the Core of Software QA Testing

Ensuring software meets quality standards necessitates rigorous software quality assurance testing, a process that typically requires manual test case development by a human tester to identify bugs and other discrepancies within the software interface.

A startup based in Raleigh, North Carolina named MuukTest unveiled an AI-driven agent on Tuesday, engineered to streamline the creation of these tests with a higher level of automation.

“The founding goal has been the automation of software QA technologies to simplify the process of test development to a mere click, encompassing the entirety of our vision,” Ivan Barajas Vargas, the co-founder and CEO, shared with TechCrunch.

These advancements empower testers to scrutinize every aspect of the software’s user interface, including menus, buttons, and operations across varied scenarios, to uncover as many defects as possible prior to the software being launched to end-users.

Renan Ugalde, the CTO and co-founder alongside Barajas Vargas, who has spent two decades in software quality assurance testing, aims to leverage Ugalde’s technical expertise to encapsulate that profound knowledge into training an AI agent capable of generating comprehensive test suites.

They have harnessed an array of AI technologies for this purpose, incorporating large language models, conventional machine learning techniques, and computer vision and image recognition. “We’ve trained AI agents to emulate the reasoning of a QA tester, to comprehend the application context, recognize various interface elements, and anticipate outcomes,” Ugalde explained.

Achieving this has necessitated the use of reinforcement learning and an in-depth understanding of the broader context, distilled from the founders’ extensive experience in QA, to inform the development of the AI agent.

Over time, the term “AI agent” has come to denote software powered by AI that assists with specific tasks, yet it still lacks a universally accepted definition. For MuukTest, this agent serves as an intelligent aide, undertaking some of the repetitive tasks traditionally performed by human QA testers.

After moving to the U.S. from Mexico in 2011, Barajas Vargas began a new chapter at Dell, and Ugalde at IBM, with both advancing in their careers before launching MuukTest in 2019 with the mission to minimize the efforts involved in executing QA tests.

Initially, their solution leveraged no-code platforms and algorithms to craft these tests, but with their latest introduction of generative AI technology, customers can now simply outline their desired test suite and have MuukTest automatically generate and prepare these tests for execution with a singleton command, significantly easing the required workload, according to Barajas Vargas.

Before integrating AI into its offerings, MuukTest was already witnessing a promising market fit early last year. Demonstrating a 15x increase in revenue over the previous year, the company anticipates even more robust growth with its enhanced capabilities.

Since its inception, which saw the firm joining the Massachusetts-based startup incubator Mass Challenge, MuukTest has secured a cumulative $6 million from investments and grants. Today, with 36 full-time employees and 10 contractors, Barajas Vargas underscores the company’s commitment to prudent financial management.

The launch of the new AI agent feature is effective immediately.