Jithin Vijayadas

• Built scalable GenAI solutions using AWS Bedrock, Claude, OpenSearch, and multi-agent RAG architectures,
significantly improving response accuracy and system performance.
• Reduced infrastructure costs by 50% through serverless architecture, semantic caching, and auto-scaling
implementations using Lambda and ECS Fargate.
• Developed and deployed LLM-powered summarization and intelligent search solutions using GPT-4, prompt
engineering, and Chain of Density techniques, reducing document processing time and improving decision-making
efficiency across enterprise workflows.

• Led ML and deep learning initiatives across fraud detection, OCR, computer vision, and payment risk analytics with
measurable improvements in accuracy and compliance.
• Designed

• Engineered a dataset of over 1 million YouTube comments for sentiment analysis, greatly improving insights into
audience engagement and enhancing model training for business analysis.
• Achieved a 15% improvement in comment classification accuracy using TF-IDF weighted Word2Vec and Artificial
Neural Networks, driving better-informed business decisions.