At Ethara AI, I evaluate multimodal AI model outputs across image grounding, entity tagging, image-to-image editing, and multimodal agent tasks using structured rubrics.
I correct and refine predictions, including bounding boxes, points, and entity references, to create ground-truth data and identify model failure patterns. I also use Python to support evaluation workflows and collaborate with annotators, reviewers, and QA on evaluation quality.
I developed 10+ Python scripts as a Python Developer Intern at Codules Technologies Pvt. Ltd. In my projects, I built a resume analyzer using NLP and developed a machine learning system to assess mental health using behavioral and demographic data.
