Kyndryl has unveiled the Agentic AI Framework, a new orchestration platform specifically designed to assist enterprises in managing the increasing prevalence of AI agents. This platform enables organizations to deploy and oversee autonomous, self-learning agents across various business processes, whether they are utilizing on-premises, cloud, or hybrid IT environments.
These specialized agents have multiple roles, such as conducting data analysis, compliance checks, and resolving service desk tickets. According to Ismail Amla, senior vice president of Kyndryl Consult, as agents interact with data and outcomes, they become more proficient at decision-making, allowing them to adapt workflows automatically. The orchestration engine behind this framework is crucial for interpreting this data, enabling enterprise systems to respond to evolving conditions in real-time.
The framework outlines the capabilities of these agents, effectively establishing policies and boundaries for their actions across the organization. Amla emphasized that the model employs a graph-based orchestration system that fosters speed, resilience, and continuous learning. This design means the agents not only react to assigned tasks but can evolve with customer business needs, providing adaptive intelligence that scales and rationalizes their decisions.
Furthermore, the Agentic AI Framework prioritizes security by incorporating encryption and zero trust authentication methods to safeguard data. It also includes support through Kyndryl Bridge, a flagship service that combines management, observability, and automation tools, leveraging AI and machine learning to offer insights into IT operations. Kyndryl boasts that this Bridge platform generates more than 12 million AI-driven insights monthly.
In addition to the technology, Kyndryl offers consulting services to aid customers in pinpointing use cases for agentic automation and evaluating existing workflows and systems. As organizations navigate increasingly complex digital landscapes, the ability to effectively manage intelligent and autonomous agents is becoming vital, as traditional automation solutions struggle to keep pace.
