Eigen RadarAI
Archive
Analysis

IBM opens Bob deployments that keep development on customer infrastructure

IBM announced a self-hosted option for Bob, its AI platform for software development and modernization. Organizations can run supported configurations on their own premises, private clouds or isolated networks. Supported local models can operate without an external connection; a separate hybrid setup connects to external model services. The choice targets organizations managing sensitive source code and regulated information under their own infrastructure and access policies.

Artificial Intelligence··Night
An IT administrator works on a laptop in a company server room.

Bob gains a customer-managed deployment option

IBM announced a self-hosted deployment option for Bob, its AI platform for software development and modernization. Organizations can run supported configurations in on-premises, private-cloud or sovereign-cloud environments. The option gives enterprises a way to keep development within infrastructure they manage.[1], [2]

IBM identifies proprietary source code, regulated customer information and mission-critical infrastructure as workloads with constraints on public AI services. The company says Bob can work where the code, application context and data already reside. Customer-managed deployment places data residency, infrastructure security policies, access and development workflows under the organization’s own control.[1]

Local models can run without an external connection

Supported models can run on premises, including in networks isolated from external connections. Organizations use models they have licensed for the supported configuration. A separate hybrid arrangement connects Bob to external model services. That route is optional: an external service connection is not required by the isolated deployment. IBM limits its announcement to supported models and configurations rather than universal local-model compatibility.[1]

Modernization spans more than cloud-native software

The offering extends beyond isolated code generation to software delivery and modernization workflows. IBM describes enterprise projects that span on-premises applications, private clouds and other controlled environments. Neel Sundaresan, the company’s general manager of AI and Automation, links the deployment choice to operational control for sensitive code and regulated data. Separately licensed models and the organization’s operating policies remain part of that deployment framework.[1]

References

  1. News sourceIBMIBM opens self-hosted deployment for Bob↩1↩2↩3↩4
  2. News sourceTechNewsReelIBM opens Bob deployments that keep development on customer infrastructure↩
AI safety and regulationChip wars and the supply chain

More analysis from this desk

View all Brief analyses from the Artificial Intelligence desk →