IBM Advances Enterprise AI Software Development with Multi-Agent Capabilities and Specialized Modernization Workflows

Details

IBM has shipped a substantial update to IBM Bob, its enterprise-focused agentic development platform, addressing a problem the company is hearing broadly from large organizations: AI has already accelerated code writing, but the bottleneck has shifted downstream to review, validation, and integration. IBM cites a survey in which 85% of DevSecOps professionals agree with this framing. The headline additions are multi-agent coordination and Bobalytics. Multi-agent support lets engineering teams orchestrate multiple specialized AI agents across the full software development lifecycle rather than running isolated single-agent coding tasks. Bobalytics is a new analytics dashboard surfacing productivity, quality, performance, and AI cost data — targeting the unpredictable-spend problem that comes from running AI at scale across large engineering organizations. IBM has also added pre-built, customizable workflows for three legacy modernization scenarios: IBM Z (mainframe), IBM i with RPG codebases, and Java. Jack Henry, a financial services technology provider, is cited as a customer using IBM Bob to accelerate RPG development and extract institutional knowledge from decades-old systems. IBM positions Bob as an end-to-end agentic development partner with enterprise governance and cost controls — competing against coding assistant vendors by emphasizing lifecycle breadth rather than point-solution quality. The announcement was made by Neel Sundaresan, GM of Automation and AI at IBM. Concrete performance numbers and pricing details are absent from the release.