Janus materials are an emerging class of two-dimensional materials with a diversity of two exclusive sides, which embark on various new multifunctional properties for electronics, optoelectronics, and memory application devices. Evolving technologies like neuromorphic computing based on floating-gate transistors, architecting an advanced artificial intelligence technology (AIT) to emulate efficient brain-like synaptic functions. In this study, we present an emerging memory design using Au/hBN/WSSe and Gr/hBN/WSSe heterostructures on the same WSSe channel, where gold and graphene serve as floating-gate materials and hexagonal boron nitride (h-BN) as an effective tunneling layer. By comparing the performance metrics based on device configurations under controlled conditions, we achieved a current ON/OFF ratio (∼10 5 ) and (∼10 3 ) for Au and few layer graphene as floating gates, respectively. The memory devices with Gr floating gate demonstrated the significant and consistent memory window of ΔV = 65 V compared to Au (ΔV = 51 V). Further, Gr/hBN/WSSe showed promising endurance (10 5 cycles) and retention (10 6 s), having gate-dependent multi-states for erase and program. Moreover, we used an artificial neural network (ANN) for digit-MNIST and Fashion-MNIST simulations, which achieved 87 % and 78 % accuracy, respectively. Simulations of WSSe-based synaptic transistors further demonstrate their capability to support ANN learning, underscoring the potential of this platform to drive next-generation AIT for memory and computing systems.