At Concirrus, I owned product strategy, roadmaps, and delivery for AI-powered underwriting and document intelligence platforms. I helped improve document extraction accuracy from 70% to 98% through taxonomy optimisation, experimentation, and collaboration with ML and Software Engineers.
I designed taxonomies and extraction frameworks using 300+ real-world insurance documents, and owned automated generation of 1,000+ insurance policy documents. Translating underwriting rules into detailed requirements helped reduce manual turnaround time by 95%, while automated validation workflows reduced manual review effort by 80%.
At Holland and Barrett, I delivered CMS features for B2B wellness campaigns and supported 10+ campaign launches. Earlier, as a Software Project Engineer at Wipro Limited, I performed 100+ patch installations in UAT and led a transition to automated testing that cut test execution time and closed CI/CD process gaps.
