DrParisa Jamadi Khiabani

I am a Postdoctoral Research Associate (Generative AI Engineer) at Imperial College London, where my work focuses on Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), multimodal AI, and production-ready AI systems. My current research develops intelligent feedback assistant systems that integrate lecture materials, lab guides, and student submissions to generate context-aware, multimodal feedback through scalable AI pipelines.

My work combines Natural Language Processing (NLP), document understanding, information extraction, prompt engineering, vector databases, and model evaluation with hands-on engineering experience in backend/frontend development, cloud deployment, and high-performance computing (HPC) environments. I have developed end-to-end AI workflows using Python, PyTorch, LangChain, AWS cloud services, REST APIs, and modern ML frameworks, with experience in scalable deployment, GPU-based experimentation, system integration, and AI pipeline optimisation. I also focus on reliability, hallucination detection, prompt safety, and secure deployment of AI systems for real-world applications.

Beyond generative AI, my research spans natural language processing, computational social science, stance detection, and sentiment analysis in social media. I have developed graph-enhanced NLP models that combine transformer-based language representations with social network information to better understand how opinions and perspectives are formed and expressed online. My broader goal is to bridge advanced AI research with practical engineering solutions that deliver reliable and meaningful real-world impact.