Singtel RE:AI and WEKA to build sovereign AI infrastructure for ASEAN

Singtel RE:AI and WEKA have signed an MOU to develop sovereign AI infrastructure for governments and enterprises in Singapore and ASEAN.

RE, the sovereign AI cloud business under Singtel Digital InfraCo, has signed a memorandum of understanding with WEKA to develop sovereign AI infrastructure solutions for Singapore and the wider ASEAN region.

The agreement brings together Singtel's sovereign AI cloud infrastructure, its Paragon network orchestration platform, and managed services with WEKA's data infrastructure technology. The companies said the collaboration is aimed at supporting governments and enterprises that need AI systems aligned with data residency, regulatory, and operational requirements.

Singtel launched RE in October 2024 as an AI cloud service for enterprises and public sector customers. The service was introduced to provide access to AI infrastructure without requiring organizations to invest in or maintain in-house infrastructure and supporting resources.

RE sits within Singtel Digital InfraCo, which brings together data centers, subsea and satellite connectivity, and Paragon orchestration for networks, clouds, and AI workloads. The agreement brings WEKA's data infrastructure technology into Singtel's sovereign AI cloud offering.

AI data infrastructure

WEKA will provide NeuralMesh, its software-defined storage system for AI workloads. The system is designed to keep graphics processing units supplied with data, reducing storage bottlenecks and limiting idle GPU time.

Singtel plans to integrate NeuralMesh into its GPU-as-a-Service offering. That service is provided through a Centre of Excellence for Applied AI, which Singtel Digital InfraCo launched with NVIDIA in February 2026.

Singtel said the center combines NVIDIA platforms, models, and blueprints with its AI cloud, data center, and network capabilities. The addition of NeuralMesh is intended to provide the high-performance data layer for AI workloads running on that infrastructure.

Bill Chang, CEO of Singtel Digital InfraCo, said AI infrastructure requirements are no longer centered only on compute capacity. He said data movement to GPUs has become a key constraint, making high-speed data architecture an important part of AI infrastructure.

Chang said the partnership combines cloud infrastructure, data capabilities, and network intelligence to support AI workloads for customers operating in sovereign environments.

Singtel has also allocated capital to the infrastructure areas covered by the agreement. In its FY26 results, the company said an additional S$1.2 billion would primarily be invested in data centers, equipment, fit-outs for GPU-as-a-Service facilities, and AI.

Deployment models

Under the MOU, Singtel and WEKA will offer a Sovereign AI Factory as a managed service for governments and enterprises in Singapore and ASEAN. The service will support use cases in national digital services, financial risk management, healthcare diagnostics, public safety, transport, utilities, and related operations.

The offering will be available in three deployment models. A Dedicated Sovereign Pod will provide single-tenant infrastructure for sensitive workloads. A Regulated Multi-Tenant Sovereign Zone will offer isolated shared environments for regulated industries. A Hybrid Sovereign Model will connect on-premises systems with sovereign cloud capacity.

The companies said the models are intended to address performance, compliance, and cost requirements.

WEKA co-founder and CEO Liran Zvibel said sovereign AI in ASEAN involves more than the location of stored data. He said the infrastructure must support AI workloads across the wider ecosystem, including public and private sector use.

Singtel's role as both a connectivity provider and cloud infrastructure operator is part of the planned architecture. AI models can be developed and governed in sovereign data centers, then extended to edge locations where low-latency processing is required.

The companies identified traffic and video analytics, airport and port operations, and industrial automation as examples of edge use cases. These workloads require AI systems to process data close to where it is generated.

The platform is also designed to support AI agents used in multi-step workflows, including applications that require sustained access to large datasets.