As the business scaled, its QA lifecycle struggled to keep pace. Test case authoring remained largely manual, requiring significant effort from specialists and limiting how quickly teams could respond to new releases or configuration changes. Regression coverage was inconsistent, and access to automated testing was effectively restricted to a small group of automation engineers. Business analysts and UAT teams had no practical way to trigger or manage regression batches independently.
This created a compounding problem. The more the product and sales motion grew in complexity, the heavier the manual burden became. Scheduled releases required disproportionate QA cycles, slowing delivery confidence and increasing the risk of defects reaching production. The organization needed a more scalable, structured approach to quality engineering, one that could grow with the business rather than constrain it. The question was not whether to modernize, but how to do it without disrupting ongoing delivery cycles.
Solution
Brillio was selected as the strategic Salesforce partner, with its Summit Partner status and hands-on experience delivering Salesforce CPQ implementations giving the client confidence in both platform depth and execution maturity.
The approach centered on embedding artificial intelligence (AI) and automation across the entire QA lifecycle, not just at its edges. Using Glean, CRT, and qTest, we built an end-to-end AI-augmented testing model. AI capabilities interpreted Jira requirements directly and generated test cases from them, cutting manual authoring time significantly. Natural-language prompts then made it possible for UAT teams and non-technical business users to independently trigger curated regression batches, no automation expertise required.
Beneath the surface, large language model (LLM)-based record-and-playback capabilities enabled rapid creation of business flows through behavior-driven development (BDD)-style scripts. AI-led hotspot analysis identified the functional areas most prone to regression and triggered targeted test reruns, concentrating effort where risk was highest.
The governance layer was equally deliberate. A GitHub-based regression suite integrated with mature CI/CD pipelines provided consistent scheduled, smoke, and ad-hoc test execution across environments. Copado and CRT frameworks supported functional and BDD-style automation, while Glean served as the unified AI interface across authoring, regression, and flow creation.
Rollout followed a phased model: a maturity and gap assessment first, then hands-on enablement workshops, then a joint pilot automation phase with the client, and finally a scaled program built on reusable libraries and standardized templates. The path forward targets approximately 90% automation coverage over time.
