Rocket expands AI controls for mainframes in banks

Rocket Software has expanded its EVA agentic AI platform for mainframe systems, targeting banks and other financial institutions facing skills shortages and tighter AI governance demands.

The update adds new operational use cases for EVA and introduces Rocket PlanGuard, a security layer that places a policy checkpoint between AI reasoning and system execution. The changes are intended to help organisations apply AI to mainframe environments while keeping actions within approved boundaries and audit controls.

Rocket is targeting a market where ageing specialist workforces and regulatory scrutiny are shaping technology decisions. Research it commissioned from Hanover Research found that 81% of financial services IT leaders described the mainframe skills gap as very or extremely significant, while 87% said AI would help address it over the next two years.

The research also suggested governance is becoming a central barrier to broader AI deployment in banking and financial services. It found that 78% of IT leaders in the sector believe regulatory considerations significantly or extremely limit their ability to deploy AI, while 46% ranked strong governance controls and approval workflows as the top requirement for AI in production.

New controls

PlanGuard is designed to extend those controls by adding a policy decision point and identity controls to AI-driven actions on the mainframe. It provides auditable, tightly scoped agent access to mainframe resources, with human oversight where required.

EVA uses natural language queries to analyse operational data and automate selected tasks on core systems. The platform can correlate information across operational environments in real time, potentially reducing the amount of manual investigation needed by operations and infrastructure teams.

The expansion comes as organisations running core banking, insurance and government systems look for ways to retain institutional knowledge as experienced mainframe staff retire. Rocket argues that AI tools can help less specialised staff work with those systems by surfacing data, logs and system context through a conversational interface.

Customers are using EVA across operations, diagnostics, security, batch processing, application management and data access. Use cases include job failure pattern analysis, dashboard and KPI generation, workflow automation, end-of-month financial reporting analysis, online banking application optimisation, queue analysis, vulnerability detection, compliance tasks and patch management.

Other uses cited by Rocket include monitoring batch performance against service-level agreements, investigating the root causes of processing failures, analysing high CPU utilisation, and identifying issues in Db2 data sharing, buffer pools and cache performance. These tasks have traditionally required specialised expertise and multiple tools.

Regional footprint

Rocket has more than 100 banking, financial services and insurance customers across Asia-Pacific, including 20 in Australia. It also has two AI pilots or proof-of-concept projects under way in the region, according to figures released with the announcement.

Globally, organisations in financial services, government, insurance, retail and telecommunications are taking part in EVA pilots. The pilot structure is intended to let customers test operational use cases with their own data and move from installation to initial insights within a short timeframe.

Rocket is also extending EVA through a model-agnostic architecture and what it describes as a standards-based approach to connecting z/OS data. The aim is to simplify access to operational and enterprise data while maintaining governance controls, an issue that remains sensitive for heavily regulated sectors that continue to rely on mainframes for transaction processing and record-keeping.

Milan Shetti, President and Chief Executive Officer of Rocket Software, said the company sees a significant opportunity to bring AI into established core computing environments without losing operational discipline.

"Enterprises have relied on the mainframe to run their mission-critical workloads for decades," said Shetti, President and Chief Executive Officer of Rocket Software. "AI promises to unlock even more value from these environments, but only if enterprises can deploy it securely and without disruption. We're helping customers apply agentic AI to mission-critical systems with speed, confidence, and control, closing the skills gap and putting that expertise within reach of every enterprise team."