At GlobalLogic Technologies Limited, I evaluate LLM responses in Google’s AI Overview for relevance, accuracy, and hallucinations, processing 32 queries a day to support RLHF fine-tuning pipelines. I categorize reasoning errors and retrieval failures, and maintain 95%+ SLA adherence while collaborating on GenAI evaluation standards.
I built an AI-powered business analytics dashboard using LangChain and LLM APIs to turn natural-language questions into SQL queries, achieving 90%+ accuracy. In an internship at Cacta, I developed scraping pipelines, RESTful APIs, and LLM-powered data extraction tools that reduced manual effort by approximately 60%.

