Architecting Scalable Intelligence for High-Throughput Autonomous Applications through Generative AI Integration with Systems Programming and Cloud Microservices

Abstract

The rapid expansion of autonomous systems in high-throughput environments has intensified the demand for intelligent, scalable, and resilient software architectures. This study presents a novel framework that integrates Generative Artificial Intelligence (GenAI) with systems programming and cloud microservices to architect scalable intelligence for next-generation autonomous applications. The proposed architecture is structured around three synergistic layers: GenAI-enhanced decision engines, performance-optimized systems code (developed in Rust and C++), and cloud-native microservices orchestrated via Kubernetes. Use cases involving real-time logistics and smart-city surveillance were developed to benchmark the framework under varying operational loads. Results indicate a substantial improvement in decision accuracy (up to 97.8%), a marked reduction in CPU and memory usage (up to 38% and 60% respectively), and robust system uptime (>99.98%) across stress scenarios. Statistical analyses confirm the significance of these performance gains. Furthermore, latency distributions and autoscaling responses reveal the architectural readiness for dynamic, distributed deployments. This study establishes a future-oriented blueprint for architecting intelligent, high-throughput autonomous systems that are both resource-efficient and operationally resilient.