The rapid increase in demand for cloud computing devices and advancements in AI services within compute continuum and high performance computing environments have introduced new challenges in integrating and optimizing device, edge, and cloud resources in modern computing ecosystems. The Integrated AI Scheduler (IAIS) presents an advanced solution for optimizing job scheduling in Kubernetes-based HPC and cloud environments. By employing AI models such as Recurrent Neural Networks (RNN) and Temporal Convolutional Networks (TCN), the scheduler dynamically learns which scheduling strategies optimally adjust resource allocation based on real-time and historical data. Additionally, the integration of a fuzzy access controller, which configures the access of jobs to the resources on different nodes, and vector databases to enhance similarity search and decision-making allows the system to optimize workload distribution efficiently and learn from prior scheduling decisions based on system performance metrics. The IAIS optimizes throughput, minimizes latency, and enhances resource utilization by using Proximal Policy Optimization (PPO) for dynamic scheduling, selected after testing RNNs and TCNs, combined with efficient hashing and caching mechanisms, as evaluated through the DECICE framework.