Adem Zarai

At FLESK, I processed and analyzed structured and unstructured data to support application features, working with APIs and data flows to maintain data consistency and accuracy.

For FleskTask, my final-year project, I designed NLP features for task analysis, including classification and prediction workflows. I also built prompt-based systems simulating LLM training and evaluation scenarios.

I handled JSON and CSV datasets and integrated them into machine-learning pipelines, refining prompts to improve output accuracy. My work focuses on text classification, sentiment analysis, and entity recognition across Arabic and English.

I'm trilingual in Arabic, English, and French, and I'm remote-ready with flexible hours. I'm seeking to contribute to AI model training and evaluation.