When it comes to the dissemination of information and ease of receiving answers to queries about dyslexia, the prevalence of a dyslexia-specific chatbot is sparse. Our effort here is to create a chatbot model that anyone with queries about dyslexia can use. In this study, we propose a Reinforcement Learning based Seq2Seq model, merged with K-Fold cross-validation. Using the BLEU score as the reward mechanism, iterative updates are made to the Reinforcement Learning parameters so that the model learns to adjust its predictions based on the rewards it receives. We have also created a dyslexia-centric conversational dataset with the help of some reliable online resources like the British Dyslexia Association and the International Dyslexia Association. The diversity and coverage scores of our dataset are 0.8541 and 0.9172, respectively. Our proposed model could achieve an accuracy of 0.9678, an F1 score of 0.9570, and a BLEU score of 0.9473.