AI-POWERED AUTONOMOUS MICROSERVICES: A SELF-HEALING APPROACH USING MACHINE LEARNING TECHNIQUES

Purpose: This paper investigates how AI, especially machine learning (ML), can enable self-healing capabilities in microservices architectures, enhancing system resilience and reducing downtime.

Design/methodology/approach: The study adopts a conceptual-analytical approach, supported by a literature review of 10 foundational papers published before 2020. It presents frameworks for self-healing through intelligent monitoring and autonomous decision-making using supervised and reinforcement learning. Findings: Self-healing microservices can autonomously detect, predict, and resolve failures using AI models trained on telemetry, logs, and behavioral patterns. These models outperform rule-based systems in adaptability, especially in cloud-native environments. Practical implications: This approach reduces system maintenance costs, minimizes human intervention, and enhances availability. It is particularly beneficial for large-scale distributed systems with dynamic workloads. Originality/value: The fusion of ML with microservices self-healing remains underexplored. This work contributes a structured perspective and visual frameworks for implementation, offering value to system architects and developers.