How Bombay Stock Exchange ( BSE ) Optimized Testing Time by 90%
A pillar of the Indian economic landscape, with over 5500 listed companies, the 141-year-old exchange is the largest in the world. It’s also the fastest stock exchange in the world with an estimated 400 million orders per day trading at a median speed of 6 microseconds.
Challenges on the way to becoming the fastest stock exchange in the world: 900+ brokers at over 100,000 branches, in addition to millions of individual traders place around 400 million orders every day on the BSE Online Trading Platform (BOLT Plus) as well as BOLT on Web (BOW). The sheer volume of transactions requires seamless, error free performance.
The transactions being carried out were high value and critical in nature, robust operation became equally critical. Everything from legacy systems running proprietary software, to desktop-based systems and web apps were actively engaged in the overall eco-system. Complexity of the QA requirement matched the complexity of systems at play. Frequent software release cycles, heavily influenced by SEBI regulations meant that the smallest of error could
impact financials and credibility. Heavy reliance on manual testing through a large Indian IT firm affected coverage and response time.
Qualitia – a script-less test automation platform for Selenium/ Web Driver was used. The easy, script-less approach to creating the test automation suites allowed the existing testing teams and the subject / domain matter experts to contribute. Test automation suites were created without significant inputs from automation experts. The execution of the tests could be carried out in unattended mode and automation experts could come into play to analyze test results post execution. A window into business processes was opened, allowing all business flows to be incorporated into the suites. Test cases were optimized to remove redundancies and duplications. Multiple Segments & Session-wise executions were carried out across different environments over a period of time.
Outcomes and Impacts
- Greater coverage of code under test enhanced reliability and prevented errors
from reaching end users
- Finding defects early in the cycle helped reduce the cycle time to fix them
- ‘Go to Market’ cycle time was reduced in line with business imperatives
- Manual efforts were reduced by 51%
Projected team size was reduced by 41%
- Knowledge and experience of existing manual testers could be leveraged
- Subject matter and domain experts could contribute without needing any automation upscaling
- Test cases were optimized to remove redundancies
- Enhanced reliability and consistency in execution results was experienced
- Increased automation helped made regression testing more rigorous