Enterprise computing faces major challenges such as ever-increasing size and complexity and escalating costs of administering mission-critical IT systems, numerous human-caused failures and outages of IT systems, lack of sufficient supply of trained system administrators, and major obstacles in dealing with changes throughout the software system life cycle. Thus, it is practically impossible to rely on human intervention and administration for development, integration, deployment, and management of enterprise IT systems in mission critical applications. This has led to the development of autonomic computing that is derived from human body?s self-regulatory nervous system. Autonomic computing is an emerging trend for building nextgeneration enterprise IT systems. Autonomic systems are envisioned to be self-aware and able to selfmanage. The focus of our research is to develop multi-agent autonomic architectures. The application domains range from telehealth and telemedicine to quality management and space exploration. This paper presents our multi-agent autonomic architecture design.