Sarcouncil Journal of Engineering and Computer Sciences

Abstract: The rapid evolution of distributed computing necessitates resilient and scalable system architectures capable of handling dynamic workloads and mitigating failures. This study explores the role of Artificial Intelligence (AI) and software engineering methodologies in optimizing distributed systems for enhanced fault tolerance, load balancing, and scalability. AI-driven approaches, including reinforcement learning-based load balancing, predictive failure detection using LSTM models, and self-healing mechanisms, were integrated into a microservices-based distributed system architecture. Experimental evaluations demonstrated a 50% reduction in latency, a 60% improvement in throughput, and an 85% decrease in failure rates compared to traditional methods. AI-based failure prediction models, particularly LSTM, achieved a 94.8% accuracy rate, significantly reducing system downtime. ANOVA statistical analysis confirmed the high significance of AI interventions (p < 0.005) in optimizing system performance. Furthermore, scalability tests showed AI-enhanced systems efficiently managed 30,000 requests/sec with controlled CPU and memory utilization. These findings establish AI as an essential component in modern distributed system design, ensuring higher efficiency, reliability, and business continuity. Future research should explore hybrid AI-cloud frameworks for further advancements in self-optimizing distributed systems