Fish farming is a cornerstone of livelihoods across Sub-Saharan Africa, yet smallholder farmers often lack timely access to expert knowledge and decision-support systems. Cage aquaculture, a growing sector in sustainable food production, faces critical challenges such as water quality degradation, disease outbreaks, and greenhouse gas emissions. Traditional machine learning approaches, which rely on extensive datasets, struggle in data-scarce environments, limiting their effectiveness in supporting aquaculture resilience and adaptation. This study proposes an AI-driven Question-and-Answering (Q&A) algorithm that leverages Few-Shot Learning (FSL) to bridge the aquaculture knowledge gap. Unlike conventional fine-tuning methods that suffer performance declines with limited data, prompt-based fine-tuning techniques, such as FewshotQA and Null Prompting, enable models to generalize effectively from minimal examples. While these approaches have been explored in text classification, their application in domain-specific extractive Q&A for aquaculture remains underexplored.