PPT Slides to Answers: Retrieval-Augmented Generation Question Answering Systems

RAG has emerged as a key technique for enhancing LLMs by reducing hallucinations and incorporating external knowledge, making it particularly valuable for domain expert QA. However, developing such systems in low-resource settings presents several challenges: (1) handling heterogeneous and unstructured data sources, (2) optimizing retrieval phase for reliable answers, and (3) evaluating generated answers across diverse aspects. To address these, we introduce a data generation pipeline that transforms raw multi-modal data into structured corpus and Q\&A pairs, an advanced re-ranking phase improving retrieval precision, and a reference matching algorithm enhancing answer traceability. Applied to the automotive engineering domain, our system improves factual correctness (+1.94), informativeness (+1.16), and helpfulness (+1.67) over a non-RAG baseline. These results highlight the effectiveness of our approach in building reliable, domain expert RAG systems with strong answer grounding and transparency.