ADDA SIVA PRASAD

At Innodata Inc., I evaluated large language model outputs across 500+ daily records, maintaining above-baseline accuracy on quality audits. I also reviewed AI-generated content for accuracy and relevance through text classification and sentiment analysis. When I found ambiguous cases or inconsistent model outputs, I escalated them to QA teams and resolved 95%+ of issues within SLA. I adapted to evolving evaluation guidelines and shared daily written status reports. At Accenture, I validated and maintained vendor master data in SAP MM, achieving error-free capture on 200+ records per sprint. I also prepared status updates and handoff documentation across three release cycles. For my AI-Based Resume Screening System project, I trained a classification model on 1,000+ labeled resumes, cutting manual review time by 70%. I built a Streamlit interface so non-technical HR staff could upload and screen resumes without engineering support.