Solution architecture
This section discusses the solution architecture for the SaaS deployment of IBM® Bob Premium Package for Z. It outlines the key components, their interactions, and how architects and developers leverage the system to analyze, refactor, and modernize mainframe applications.
The IBM Bob Premium Package for Z extends the capabilities of IBM Bob IDE (agentic experience) and Bob Shell by integrating:
- Z application understanding services
- Refactoring assistance
- Workflow-driven automation
- Cloud-based AI services and large language models (LLMs)
IBM Bob Premium Package for Z on SaaS
The architecture is hybrid and consists of three major domains:
- User Interaction Layer (IBM Bob IDE or Bob Shell)
- Z Understand (Linux® VM s390x or x86)
- Cloud AI Services Layer (IBM Bob in Cloud)
The IBM Bob Premium Package for Z delivers a role-based experience tailored to both architects and mainframe developers, enabling them to interact with mainframe applications through a unified, agentic interface in Bob IDE.
Architect
The solution focuses on application understanding and strategic analysis. Working in the Z Architect mode, the architect initiates analysis workflows that leverage the IBM Bob Premium Package for Z extension to orchestrate backend services. These workflows invoke Z Understand services running on a Linux VM (s390x or x86), where the system builds an analysis project and associated database that captures detailed metadata about application structure, dependencies, and data flows. The output is a comprehensive knowledge model of the application, along with generated documentation and visualization artifacts, which are presented back in the IDE. This enables architects to rapidly understand complex systems and make informed modernization decisions with significantly reduced manual effort.
Mainframe Developer
The solution helps developers improve productivity in coding, explaining, and refactoring programs. Working in the Z Code mode within the same IDE, developers use natural language or commands to interact with the system. The IBM Bob Premium Package for Z extension turns these inputs into workflows and reusable skills that work across both on‑premises and cloud systems.
Z Understand analyzes the code and provides context, while refactoring services help update and improve it. Requests are sent to IBM Bob’s cloud services, where AI models generate outputs like explanations, refactoring suggestions, and documentation. These results appear directly in the IDE, helping developers understand legacy code, improve quality, and work faster without switching tools.
The solution uses a unified architecture with shared components for consistency and scalability. The extension simplifies complex tasks, the on‑premises layer keeps sensitive data local, and the cloud layer provides advanced AI capabilities. Together, this setup enables architects to focus on system design and developers to focus on coding, creating a smooth and efficient development experience for IBM Z.
Large Language Models
Large language models (LLMs) are integrated through the IBM Bob cloud backend, which connects to managed LLM services (such as AWS Bedrock) to process user requests and generate responses.
When an architect analyzes an application or a developer requests code explanations, generation, or refactoring, IBM Bob automatically selects the most appropriate language model for each task based on complexity, required capabilities, and cost-efficiency. This intelligent routing ensures you receive high-quality responses while optimizing resource usage.