Manual testing is rapidly making way for automated testing as more and more banks, financial services firms and other players seek for ways to best more frequently and, more importantly, more thoroughly and accurately.
However, many existing approaches, although valuable, are often seen as reactive and merely identify issues and flaws after they have occurred. This can lead to mounting costs and significant reputational damage. The recent CrowdStrike/Microsoft saga is a good example of that.
Firms should therefore opt for another strategy, see the bug coming before it’s there, basically. In other words: predictive testing.
At least, that is the message from Maneesh Sharma, since May 2023 the COO of AI-powered test execution cloud platform LambdaTest.
He found that adopting a predictive testing approach with AI can help firms identify and mitigate issues earlier in the software development lifecycle.
“I have seen how earlier detection can help lower costs, ensure a faster time to market and enhance software quality,” Sharma said.

In financial services and fintech, predictive testing can identify anomalous transactions that may trigger fraudulent activity, thus preventing such transactions in real time.
“Within fintech apps, it can be used to simulate potential cyberattacks and identify vulnerabilities so developers can fix these issues in the architecture before attackers can exploit them,” he said.
Overall, Sharma sees software testing is moving from a reactive approach to a predictive approach.
“With the help of machine learning and vast amounts of testing data, I foresee QA teams becoming more like strategic advisors instead of just fixing bugs,” he observed.
“Predictive software testing is all about leveraging data insights to build more sophisticated software that avoids failures and is released faster,” he argued.
By leveraging AI/ML algorithms, predictive testing can analyse vast data sets, including past test results, code structures and user behaviour patterns.
“By identifying trends and correlations, these algorithms can help predict potential issues,” he wrote in a recent Forbes analysis.
He singled out a recent study by computer giant IBM, which found that AI-based testing can reduce software defects by up to 30%.
“AI can further anticipate the specific functionality that might crash under heavy load by analysing millions of lines of code and user interactions,” Sharma added.
Not unimportant, predictive testing can also help uncover edge cases, leading to more comprehensive testing.
Further, “by analysing the historical test execution data for patterns and unusual behaviour, predictive testing can identify and flag flaky tests,” Sharma said.
As a part of this, firms can also measure your flaky test debt, allowing developers and quality analysts to refine the user interface and user flow before real users stumble, leading to a more intuitive and enjoyable experience.
Challenges
While there are many benefits promised by predictive testing, Sharma is the first to acknowledge that implementing this technology requires overcoming certain challenges as he stressed QA teams can encounter some major obstacles while integrating it into their real-world workflows.
Firstly, the biggest challenge with modern-day quality engineering is consolidating test data from multiple sources.
“A simple checkout flow would involve multiple applications programming interfaces (APIs), both internal and external to complete the test build,” Sharma noted.
“Debugging test failures with fragmented test data becomes nightmarish in complex workflows, making it difficult to pinpoint what led to test failure and where,” he stressed.
To streamline the process, firms can start by implementing robust data governance practices and centralising the data aggregation process.
“Predictive software testing is all about leveraging data insights to build software that avoids failures and is released faster.”
– Maneesh Sharma
Toward this, having a unified test execution platform can help, Sharma argued. “Gather data from multiple sources and highlight test execution data in one place.”
He found that this enhances test observability by grouping and classifying errors, highlighting error trends and identifying flaky test percentages in your overall test build.
Another major bottleneck may be overcoming resource resistance and the skills gap.
According to research by Sharma’s firm, more than half of all firms said there is a lack of skilled professionals in the field of AI.
Furthermore, resistance to AI remains high with many mid-sized companies hesitant to adopt AI-based testing due to factors like potential job losses and complexities.
“To overcome the skill gap, leaders should encourage knowledge transfer through regular senior-led training and build collaboration with joint Dev/QA code review sessions,” Sharma said.
“They can also ensure that their testing platform includes methods to add comprehensive documentation, tutorials and community forums,” he noted.
To overcome resource resistance to AI, Sharma recommends that you showcase initial AI success on smaller projects to demonstrate its perks, “be it in the form of cost reduction or efficiency gains.”
He added: “Once the gains are evident, resources should feel more aligned to incorporate and expand AI in existing testing workflows.”
UPCOMING QA FINANCIAL EVENTS

READ MORE
- Kobiton launches new capabilities for mobile testing
- ECB names new digital chief as QA climbs the priority ladder
- Maveric eyes IPO amid major US growth plans
- Exclusive: ‘AI productivity means more risk,’ warns Tricentis ML head
- QualityKiosk wins major testing deal with Dubai’s largest bank
Become a QA Financial subscriber – for FREE
News and interviews * Receive our weekly newsletter * Get priority invitations to our Forum events