NVIDIA is expanding the deployment of its artificial intelligence technology across Southeast Asia. This regional growth encompasses public-sector partnerships, locally adapted foundation models and additional computing infrastructure.
These regional developments include Sea Limited becoming the first enterprise in ASEAN to adopt the NVIDIA Vera Rubin platform. Alongside Sea, an array of organisations across Singapore, Malaysia, Thailand and Vietnam are deploying NVIDIA models and software suites to support public safety, legal services, healthcare, transport and other essential applications.
Sea adopts Vera Rubin across its businesses
Sea plans to use NVIDIA Vera Rubin to develop and deploy artificial intelligence models and software agents across Shopee, Monee and Garena. Founded in Singapore, the technology company currently serves hundreds of millions of users through the three platforms.
On the e-commerce side, Shopee already uses artificial intelligence to assist sellers in creating product listings, improving product discovery and actively engaging customers. Sea also applies the technology to guide its broader business decisions. NVIDIA highlighted these features as a crucial resource for smaller businesses with limited budgets, granting them access to advanced computational tools that were previously concentrated only within larger companies.
At Monee, Sea’s digital financial services business, the technology assists with fraud detection and credit risk assessment, whilst Garena applies it within gaming experiences and online services. Vera Rubin will supply the infrastructure needed to develop and run these systems at a larger scale.
Organisations adapt Nemotron for local languages and industries
NVIDIA is also widening the use of its Nemotron open AI models, datasets and development tools throughout the region. The resources assist organisations building specialised systems for Southeast Asian languages and industry-specific applications.
The Nemotron architecture allows organisations to customise and retain direct control over models according to their own operational requirements. Alongside the models, NVIDIA is distributing persona datasets containing synthetic data that reflects local populations, regional languages and domestic workforces.
In Singapore, AI Singapore is adding Nemotron models and NVIDIA NeMo tools to SEA-LION, its open family of models designed for Southeast Asian languages and cultural contexts. Regional biotechnology is also progressing through local collaborations. Hummingbird Bioscience and LynxKite are utilising Nemotron 3.5 Lightning and NeMo Retriever to build a Toxicity Knowledge Graph. This platform combines public repositories with proprietary biological data to support drug discovery and development.
In Vietnam, Viettel AI is fine-tuning Nemotron 3 Super for local-language systems and autonomous agents that carry out multi-step tasks. The model ranked highest on both the VMLU benchmark and Viettel AI’s internal product benchmark. Viettel plans to deploy this system within Legal AI, an assistant for staff and external customers.
In addition, FPT Smart Cloud has co-developed Nemotron-Personas-Vietnam. This open dataset is based on Vietnamese demographic and cultural data, giving developers a structured resource to train and evaluate localised AI models.
In Thailand, iApp Technology is adapting Nemotron 3 Nano for legal applications as part of the ThaiLLM Collaboration. Its OpenThai 2.0 Legal model is being fine-tuned with Thai legal data and released as open source. The framework also powers Thanoy, a specialised legal-assistant chatbot that currently serves about 43,000 users.
Public-sector projects expand across the region
Government and infrastructure projects form another central part of NVIDIA’s regional activity. Across Southeast Asia, these initiatives encompass varied public-sector workloads, although their individual stages of development vary.
In Singapore, HTX, the Home Team Science and Technology Agency, is beginning research with Nemotron 3 Super and Nemotron 3 Nano Omni for public-safety applications. This technical initiative explores systems capable of handling complex reasoning under demanding operational conditions. The software is designed to combine information from vision, audio and language simultaneously. NVIDIA has described these models as a potential foundation for locally controlled artificial intelligence within Singapore’s Home Team.
NCS is using Nemotron models alongside video search and summarisation technology across enterprise and public-sector projects, alongside work on humanoid robotics. In parallel, ST Engineering deploys NVIDIA NeMo tools and cuOpt software within its AI Studio platform for marine maintenance, repair and overhaul operations.
Elsewhere across the region, public-sector initiatives are progressing in multiple markets. Malaysia’s YTL AI Labs is adapting Nemotron to support both enterprise functions and citizen services. Meanwhile, transport networks are seeing operational deployments. ITMAX is using NVIDIA Cosmos models and video-search technology for traffic operations, while Thailand’s AS-TECH is applying the same technologies to track passenger flow at airports.
