GenAI Developer (RAG & Retrieval / Hybrid Retrieval Agentic RAG) at GenTA
Developed and optimized AI image-to-video and image transformation pipelines using evaluation and performance measurement as part of an AI workflow. Implemented inference and model optimization steps to reduce VRAM usage while maintaining output quality. Contributed to deployment workflows, integration, monitoring, and performance improvements to support trained model utilization. • Designed image-to-video/image transformation pipeline components. • Reduced inference VRAM requirements from 40GB to 16–17GB via optimization. • Supported deployment, integration, monitoring, and performance tuning. • Worked within an RAG and retrieval-enhanced AI workflow context.