Deep Learning has evolved and has contributed to significant breakthroughs in fields like Natural language Processing and Computer Vision. It’s limitations are it’s inability to explain informed decisions, generalizing beyond the training data set and add in domain knowledge. Neuro-symbolic AI comes up with a very favourable solution, merging the strengths of both the archetypes that will help to create systems that have the ability to explain and reason themselves logically instead of solely being data-driven. This paper includes the core approaches, behind SynapseAI including hybrid models that incorporate and integrate the neural networks with frameworks of symbolic reasoning. We explore various architectures like neural-symbolic integration, differentiable programming and neuro-symbolic reinforcement learning. Moving ahead, we also explore it’s practical and real-world applications in sectors like healthcare, scientific discoveries and autonomous systems where constant learning and reasoning are both crucial. Despite the potential SynapseAI faces challenges in scalability, efficiency and smooth integration of Deep learning models with Symbolic knowledge. We observe and analyze the challenges and propose ways to make AI more adaptive and robust. By integrating knowledge and learning. Synapse AI can be a significant step towards new generation of AI Systems that will have the ability to learn, explain and generalize similar to humans.