How the workflow operates
The process begins with a modernization request, such as analyzing a legacy application, creating a cloud migration strategy, modernizing application code, or preparing enterprise data for AI. The Orchestrator Agent assigns activities to the relevant domain agents, which collaborate across the workflow rather than relying on a single AI agent or a purely sequential process.
As agents execute, they use Amazon Bedrock AgentCore Runtime for AI workloads and AgentCore Gateway to access the enterprise systems and services required for the task. Transformation artifacts can include migration strategies, architecture blueprints, generated code, documentation, and deployment assets. These outputs are refined through the workflow, with Amazon Aurora persisting workflow state and metadata while generated code and assets flow into Git-based delivery pipelines, and are returned to engineering teams for implementation.
Keeping people in control
Agentic transformation does not remove human accountability. Human-in-the-Loop validation enables subject matter experts to review recommendations before critical decisions are finalized, while Human-on-the-Loop governance provides ongoing visibility and oversight. Guardrails and metrics add another layer of control and observability across the transformation lifecycle.
This combination allows organizations to use AI to accelerate repetitive and complex modernization activities while preserving the transparency, governance, and expert involvement required for enterprise-scale programs.
One platform across four modernization domains
The same agentic foundation supports four major transformation domains: mainframe modernization, cloud migration, application modernization, and data migration. Specialized agents can analyze legacy environments, assess cloud readiness, generate modernization blueprints and code, create migration strategies, and prepare data for AI-powered applications.
By combining Amdocs’ enterprise modernization expertise and agentic orchestration with AWS services for running, connecting, and scaling AI agents, the joint solution moves beyond isolated automation toward coordinated, enterprise-scale transformation. Multiple specialized agents can collaborate across complex programs while human experts retain control over the decisions that matter most.
Conclusion
As enterprise technology landscapes become more complex, transformation increasingly depends on coordinating work across applications, infrastructure, data, and teams. The Amdocs Agentic Services Platform with AWS provides a model for doing that through multi-agent collaboration, serverless agent execution on AWS, enterprise connectivity, and built-in human governance—helping organizations accelerate modernization without giving up transparency or control.
To learn more about the Amdocs Agentic Services Platform and how it works with AWS to accelerate enterprise transformation, visit the Amdocs Cloud solutions page. Additional AWS resources: Amazon Bedrock AgentCore Runtime (serverless agent hosting), Amazon Bedrock AgentCore Gateway (secure tool and API connectivity for agents), Amazon EKS (orchestration layer), Amazon Aurora (workflow state), and AWS Transform (agentic mainframe, VMware, and .NET modernization).
