At UST, I designed and optimized prompts for enterprise AI use cases, including cybersecurity applications for a U.S.-based banking client. The work helped reduce manual workloads by 40%, and I developed two centralized prompt libraries to support reuse and consistent outputs.
I built and deployed two AI-powered cybersecurity agents with voice conversation capabilities into production, automating information gathering and reducing manual research time by 50%. I also evaluated the agents for bias, misinformation, hallucinations, and policy issues using safety rubrics, LLM-as-a-judge, and human review.
Earlier, I worked as an IAM Analyst at SAT Microsystems, a Systems Engineer at Zensar Technologies, and a Support Analyst at HCL Technologies. I also supported technical onboarding and knowledge transfer for new joiners at UST.
