IBM has expanded its Bob software development platform with new multi-agent capabilities, built-in AI usage and cost analytics, and pre-built workflows for modernising enterprise systems.
The updates cover software development tasks including code review, validation, legacy application updates, and large-scale modernisation projects.
AI development shifts beyond code generation
According to 2026 GitLab research cited by IBM, 85% of surveyed DevSecOps professionals agreed that AI has shifted the software development bottleneck from writing code to reviewing and validating it. IBM said Bob is designed to support work across the software development lifecycle rather than operate solely as a coding assistant.
A 2026 longitudinal preprint involving professional software engineers found that 82% of respondents reported spending less time writing code when using AI coding assistants, while the researchers identified a broader shift from code creation towards verification activities.
The researchers described the emerging work as “supervisory engineering,” encompassing the direction, evaluation, and correction of AI-generated output.
The platform can assign different models to specific tasks, coordinate execution across multiple AI agents, and track factors including productivity, quality, performance, and cost.
IBM has also introduced Bobalytics, an analytics feature that gives users visibility into AI consumption and resource allocation. Bobalytics provides information on AI consumption, resource allocation, usage, and spending.
IBM said engineering teams can face inconsistent results and unpredictable spending when developers manually choose models for different tasks and attempt to balance performance against cost. The company said Bob matches models to tasks and coordinates work across multiple agents at the execution level.
Neel Sundaresan, general manager of automation and AI at IBM, said enterprise requirements are moving beyond standalone coding assistants.
“The bar for enterprise AI is no longer a better coding assistant,” he said. “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.”
How subagents manage context
Another update adds parallel, model-native tool calling. Models can request several tools during a single turn and run them concurrently rather than invoking each tool sequentially.
Bob also now uses subagents to handle complex tasks in isolated contexts. Activities such as reading files, running searches, and tracing functions can increase the amount of information held within an AI model’s context window, adding to the volume of tokens processed.
During long, multi-stage tasks, file contents, searches, function traces, and other intermediate results can accumulate in the main context.
Anthropic’s technical documentation also describes subagents operating in separate context windows and returning results to the main conversation. IBM said Bob’s subagents handle complex work in isolated contexts to support faster responses while helping manage cost.
Modernising legacy systems
The company is also extending Bob into enterprise application modernisation through premium packages for IBM Z, IBM i, and Java environments.
The packages provide pre-built workflows that organisations can customise and extend for their own systems. IBM said the workflows are structured to produce consistent and auditable results regardless of which engineer runs them.
Long-lived codebases can contain years of architectural decisions, dependencies, undocumented knowledge, and operational constraints.
A 2026 research preprint on AI-assisted software engineering argues that source code alone does not fully capture the design rationale, contextual decisions, and historical information needed to understand long-lived software systems.
IBM’s premium packages use structured workflows designed for specific environments, including IBM Z, IBM i, and Java.
The IBM Z package supports COBOL and PL/I modernisation, as well as Job Control Language analysis. Mainframe systems remain in use across sectors including banking, insurance, and commerce.
The package applies AI-assisted modernisation workflows to applications running in IBM Z environments, including analysis of JCL, which is used to define and control jobs on IBM mainframe systems.
The IBM i package includes remote file system integration, IBM i-specific operating modes and tools, and workflows built around development and maintenance practices in IBM i environments.
A separate Java Modernization package supports migration to Java 25, large-scale code refactoring, and dependency analysis across Java application portfolios.
IBM said the pre-built workflows are intended to reduce variability in complex, multi-stage modernisation projects. The company said the sequence of tasks can be customised for an organisation’s environment while remaining repeatable and auditable across users and projects.
Financial technology provider Jack Henry has used Bob to support development work involving its RPG codebase. The company had faced challenges maintaining and updating the codebase as its application portfolio expanded in size and complexity.
“Using IBM Bob, our developers are able to accelerate RPG development workflows, improve code quality, and gain deeper insights into decades of accumulated system knowledge,” said Kevin Sligar, chief technical architect at Jack Henry, adding that the platform had also supported efficiency in enhancement work.
Cloud services and consulting company Blue Pearl has also used Bob on a legacy modernisation programme. According to Blue Pearl Group CEO Saireshan Govender, a project initially estimated to require nine months and 14 engineers was completed in three days after the company introduced IBM Bob.
“The most powerful outcome wasn’t the speed,” Govender said. “It was the combination of operational efficiency, cost optimisation, and real-world results we could trust and build on.”
(Photo by IBM)
See also: IBM and Red Hat automate open-source vulnerability remediation
Want to dive deeper into the tools and frameworks shaping modern development? Check out the AI & Big Data Expo, taking place in Amsterdam, California, and London. Explore cutting-edge sessions on machine learning, data pipelines, and next-gen AI applications. The event is part of TechEx and co-located with other leading technology events. Click here for more information.
Developer Tech News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
