IBM upgrades Bob with multi-agent tools & analytics

IBM has updated its Bob software development platform and launched three Bob Premium Packages. The changes add multi-agent capabilities, analytics on AI use and cost, and preset workflows for modernisation work.

The update is part of a broader push to move AI in software engineering beyond code generation into review, validation, governance and system modernisation. IBM said this reflects growing pressure on engineering teams as AI tools generate more code while downstream work becomes harder to manage at scale.

Bob is an agentic software development platform designed to work across the software development lifecycle rather than within a single coding tool. The latest version adds model-native parallel tool calling, subagents that manage work in isolated contexts, and a new analytics feature called Bobalytics.

Bobalytics is intended to give organisations greater visibility into AI consumption, resource allocation and spending. It also tracks productivity, quality and performance, helping teams monitor how AI-driven development is used across projects.

IBM argued that many engineering teams still manually choose AI models and weigh cost against performance, often with inconsistent results. Bob is designed to match models to tasks and coordinate execution across several agents, while providing a central view of output and cost.

A survey cited by IBM points to a shift in where development teams face bottlenecks. According to the company, 85% of surveyed DevSecOps professionals agreed that AI has moved the constraint from writing code to reviewing and validating it.

Modernisation focus

Alongside the platform update, IBM introduced three premium packages aimed at large-scale modernisation work in IBM Z, IBM i and Java environments. The packages are built as preconfigured workflows that teams can adapt to their own systems while keeping outcomes consistent and auditable.

IBM Z remains a key environment for banks, insurers and large commercial groups, and one in which AI tools have often struggled to fit established workloads and languages. Its new package for that market covers COBOL and PL/I modernisation, as well as JCL analysis.

The IBM i package is aimed at organisations running long-established operational systems on that platform. It includes remote file system integration, IBM i-specific modes and tools, and workflows based on how those environments are typically managed.

For Java, IBM is targeting one of the largest pools of legacy application work in corporate IT. The package includes workflows for migration to Java 25, large-scale refactoring and dependency analysis.

IBM said the premium packages draw on its long history in enterprise modernisation and are intended to reduce the variability that can come with AI-assisted work on complex, multistage projects. Structured workflows, it said, can help teams produce repeatable results that are easier to audit.

That pitch reflects a wider debate in the AI software market over whether coding assistants save meaningful time once review, compliance and production checks are included. Suppliers are increasingly trying to show that AI can support entire engineering processes rather than simply helping individual developers produce code faster.

Customer examples

IBM pointed to Jack Henry as one production user of Bob. The financial services and banking technology provider has been using the tool as it manages a large RPG codebase built up over time.

"Using IBM Bob," said Kevin Sligar, Chief Technical Architect, Jack Henry. "Our developers are able to accelerate RPG development workflows, improve code quality, and gain deeper insights into decades of accumulated system knowledge while gaining efficiency in enhancement efforts."

IBM also cited Blue Pearl, a cloud solutions and consulting services company, as an example of Bob being used in a modernisation programme. According to IBM, the project involved legacy systems and had initially been expected to take months with a sizeable engineering team.

"We introduced IBM Bob to a legacy modernization program, an effort originally projected to take nine months with 14 engineers was completed in just three days," said Saireshan Govender, Group CEO, Blue Pearl. "The most powerful outcome wasn't the speed - it was the combination of operational efficiency, cost optimization, and real-world results we could trust and build on."

IBM framed the new release as a response to what large organisations now expect from AI tools in software development. The market, it said, has moved beyond demand for a better coding assistant towards systems that fit existing engineering workflows while providing governance, security and spending controls.

"Bob is the platform enterprise customers have been asking for," said Neel Sundaresan, GM, Automation and AI, IBM. "The bar for enterprise AI is no longer a better coding assistant. It's an end-to-end agentic development partner that works inside any system development teams already use, with the governance, security, and cost controls enterprises require. We built Bob to solve the problems enterprises actually have, and the updates we're announcing today are the foundation for everything that comes next."