At Innodata, I evaluate artificial intelligence and large language model outputs against project criteria, reviewing accuracy, factuality, relevance, and instruction following. I also review code and technical outputs, document findings, and analyze recurring errors.
At AISPRY, I developed Python-based machine-learning and computer-vision workflows with TensorFlow and Keras. I worked across data preparation, annotation, training, validation, and model-output review.
On my Highway Inspection and Maintenance Using Artificial Intelligence project, I built an object-detection workflow for identifying road defects with YOLOv8 and OpenCV. I also built a retrieval-augmented generation application using Python, Streamlit, and FAISS.
