Reza Dorkhah

Designed and deployed machine learning models across various client projects to improve prediction accuracy and business decisions. Fine-tuned large language models to enhance chatbot interactions and task-specific applications. Built end-to-end ML pipelines using AWS services to reduce deployment time and effort. Integrated AI systems into production ensuring scalability and robust performance. Performed data analysis and visualization on structured and unstructured datasets. Key projects include GenAI for Talent Acquisition, conversational AI chat agents for finance, healthcare, retail, multimodal integrations like voice cloning and emotion recognition, OCR and vision-language models for document validation, and web automation.

Developed data pipelines and visualizations using Python and R for national survey datasets. Contributed to the development of predictive analytics ML algorithms. Collaborated with senior data scientists to refine models and improve performance. Communicated insights to stakeholders to support data-driven decisions. Highlight project included building a semantic segmentation model for food ingredient detection using vision transformers and deploying an API for SME validation.