IBM Enhances Enterprise AI Software Development with Multi-Agent Capabilities and Dedicated Modernization Workflows

The latest version of 'IBM Bob' supports rapid development of production-ready software.

Optimizes costs across the entire AI development process, not just model selection.

Offers customizable, pre-built workflows for modernizing IBM Z, IBM i, and Java environments.

[Armonk, New York, USA – July 9, 2026 (local time)]

Today, IBM announced new capabilities for its enterprise AI development partner, 'IBM Bob.' The enhancements include multi-agent functionality, an analytics feature that visualizes AI usage and costs, and pre-built, dedicated workflows to support the modernization of enterprise systems and automate development processes.

While organizations increasingly use AI to generate large volumes of code, the bottlenecks in software development have shifted to other stages. According to research, 85% of DevSecOps professionals report that 'AI has moved the development bottleneck from coding to review and validation.' IBM Bob is designed to leverage AI across all phases of software engineering. Rather than limiting AI to a single interface for individual tasks, it provides an integrated platform enabling teams to collaborate throughout the entire software development lifecycle.

For example, engineers at Jack Henry, a provider of financial services and banking technology, faced challenges maintaining and extending a large RPG codebase as their application portfolio expanded. Kevin Sligar, Chief Technical Architect at Jack Henry, stated: 'By leveraging IBM Bob, developers have been able to streamline RPG development workflows, improve code quality, and gain deeper insights into systems built over many years. Efficiency in feature enhancements has also improved.'

Many enterprise engineers manually select AI models based on cost and performance trade-offs, leading to inconsistent results and unpredictable costs. With this release, IBM Bob now optimizes not just model selection but the entire AI execution infrastructure. By assigning optimal models to specific tasks and using sub-agents to coordinate AI processing across multiple agents, the new 'Bobalytics' feature visualizes productivity, quality, performance, and cost. This enables enterprises to scale and optimize their AI usage effectively.

Neel Sundaresan, General Manager of Automation & AI at IBM, said: 'Bob is the platform enterprise customers have been asking for. Enterprise AI requires more than just a good coding assistant. It needs an agent-based partner that operates within existing development environments and supports the entire development lifecycle with governance, security, and cost management. Bob was built to solve the real challenges enterprises face. These new capabilities lay the foundation for future evolution.'

Development teams face unique challenges when applying AI not only to code generation but also to larger, more complex tasks such as legacy application revitalization and modernization of IBM Z, IBM i, and Java environments.

Blue Pearl, a cloud solutions and consulting services company, has achieved success using IBM Bob on such complex projects. Saireshan Govender, Group CEO of Blue Pearl, said: 'After introducing IBM Bob into our legacy modernization project, work expected to take 14 engineers nine months was completed in just three days. The biggest achievement isn't just speed. We also gained improved operational efficiency, cost optimization, and reliable results in real-world operations.'

AI outputs can vary depending on workflow methods, posing significant challenges for critical, multi-phase projects. By leveraging structured, highly reproducible workflows, these variations can be minimized, enabling reliable and auditable outcomes at enterprise scale.

IBM Bob now offers pre-built workflows that teams can customize and extend to fit their environments. This ensures consistent, auditable results regardless of the user. Additionally, IBM is launching the 'IBM Bob Premium Package for Z,' 'IBM Bob Premium Package for IBM i,' and 'IBM Bob Premium Package for Java Modernization.' These workflows are built on IBM’s decades of expertise and optimize AI for enterprises undertaking large-scale modernization.

New Features in IBM Bob

Built-in Usage Visibility and Cost Optimization: The new 'Bobalytics' feature enables monitoring of AI usage, resource allocation, and operational status. Enterprises can efficiently scale AI usage in alignment with internal policies.

Parallel and Model-Native Tool Invocation: AI models can now invoke multiple tools simultaneously and execute them in parallel.

Large-Context Management via Sub-Agents: AI-driven operations such as file loading, search, and function tracking can inflate context size and increase costs. With IBM Bob, sub-agents execute complex tasks within independent contexts, enabling fast response times and effective cost control.

The latest version of IBM Bob is available for download at bob.ibm.com/download. For more details on the new features, visit https://bob.ibm.com/blog/bob-v2-release-announcement/.

'IBM Bob Premium Packages' Now Available

For many years, IBM has supported enterprise modernization in mainframe, IBM i, and Java environments. The first three premium packages now available deliver this deep expertise as AI-native workflows. Designed for enterprise environments where other tools fall short, they are structured, reproducible, and auditable.

The following premium packages are now available:

IBM Z: Mainframes are core systems supporting finance, insurance, and commerce worldwide and have historically been among the most challenging areas for AI adoption. IBM Bob modernizes COBOL and PL/I