Sofy launches SofySense, a generative AI-driven no-code mobile app testing solution

Software testing platform Sofy announced the launch of SofySense, an innovative mobile app testing solution that merges artificial intelligence (AI) and no-code automation. The latest offering enhances Sofy’s pre-existing platform and incorporates GPT-integrated, intelligent software-testing AI technology to provide quality assurance (QA) assistance. 

With the new solution, testers can concentrate on more mission-critical tasks, while Sofybot, the platform’s intelligent generative AI chatbot, offers prompt and accurate responses to specific test queries. The company claims that the new product can accelerate product release times by 95%, streamlining the software testing process for software developers.

Sofy’s CEO, Syed Hamid, said that SofySense’s AI technology is comparable to OpenAI’s ChatGPT, which can generate human-like text in seconds. 

“SofySense reduces time by generating test cases; eliminating the need for research and test design; allowing test cases to seamlessly convert to automated tests instead of investing time in writing automation code; and analyzing results [in order] to surface issues,” said Hamid. “With SofySense, we’ve enabled AI technology in a useful way, helping testers author and execute tests in seconds.”

SofySense builds upon the company’s 20-plus years of experience and shared knowledge, to create a seamless integration of AI and low-code/no-code. 

Using generative AI to streamline software development

According to Sofy, the company found that software developers spend between 20% and 50% of their time on testing, which accounts for nearly a third of overall IT spending. SofySense uses OpenAI’s large language models to provide advanced insights, analysis and assistance for an organization’s QA needs. 

With Sofy’s existing feature of converting manual tests to automated tests with one click, users can now easily generate insights from all testing activity on the platform.   

“Testers will no longer need to analyze test results, research all the possible test cases, implement feedback, and debug while trying to meet deadlines,” Syed Hamid, founder and CEO at Sofy, told VentureBeat. “Our new product enables users to enhance their testing strategy with faster test results; augment test creations; and be able to find answers when they need [them].” 

Testers can ask the integrated GPT-powered Sofybot chatbot to summarize the QA progress of their current release, highlighting device coverage and surface failures that may have occurred during testing.

Sofybot can directly read Confluence or any other technical documentation tool and generate test cases. Engineers can further use Sofy’s no-code automation feature to automate their test cases. Instead of a custom reporting tool, testers can use the generative AI bot to generate clear, easy-to-understand reports.

“We are using GPT-3.5 right now to analyze results,” explained Hamid. “By providing an application state-like checkout screen, SofySense uses a combination of natural language processing (NLP) and images to identify what can be tested based on over 30,000 app patterns we have seen.

“The only thing testers have to do is modify those tests or regenerate as the app functionality evolves instead of updating it manually.”

Quick, efficient data-driven decisions

The company believes that by integrating the technology into its daily workflow, its engineering teams would no longer need to spend countless hours analyzing failures and generating reports daily.

Hamid said that SofySense provides a modern way to quickly gather test results, augment test creations, and find answers to questions in one place. It streamlines testing processes while ensuring high-quality releases, making it a significant step towards true self-healing software. 

Self-healing is the ability of software systems to detect and remediate issues without human intervention.

“It’s more than just self-healing; it’s about contextualizing tests, running tests, analyzing and reporting. Today self-healing is used only to address false positives in a test run, whereas we are enabling organizations to significantly reduce their efforts on creating, running, executing and maintaining tests,” added Hamid.

Hamid said that the company’s future plan is to integrate non-functional testing using generative AI.

“We are currently working on how to test for accessibility without any human intervention, which includes lots of different system variations,” said Hamid. “We will also be performing security testing using Open Web Application Security Project (OWASP) guidelines in shifting security testing to the left (making changes in when, where and how to apply security best practices).”