Skip to main content
GPU Plexus

A Thai AI technology company

AI infrastructure built by Thai engineers

Accelerating a Smarter Tomorrow.

GPU Plexus delivers GPU Compute, LLM Hosting and API Services through to AI software development — with Swarm Manager for running GPUs inside your own organisation.

Our company

24+ yrs
Executive team experience
12+ yrs
Lead engineer experience
100%
Thai shareholders
5
Service areas

End-to-end AI services across five areas

We do more than rent out GPUs — we cover everything from infrastructure and models to applications in production.

  • 01

    GPU Compute

    High-performance AI compute clusters with GPU partitioning through MIG on NVIDIA B200 architecture.

  • 02

    LLM Hosting

    Deploy, tune and serve private LLMs and RAG inside your organisation, securely.

  • 03

    API Services

    Connect models and AI services through an API Gateway with scoped tokens and rate limiting.

  • 04

    Application Development

    Design and build AI software, data analytics systems and enterprise web applications.

  • 05

    Security & Governance

    Per-tenant network isolation, five-dimensional quotas, OAuth2/SSO support and a 365-day audit trail.

* NVIDIA and B200 are trademarks of NVIDIA Corporation

Expertise and track record by area

Our team has accumulated experience across AI systems, information systems and data infrastructure, summarised as the following areas of expertise.

  • Generative AI and RAG

    Retrieval systems built on generative AI and RAG, delivered through microservices and an API Gateway.

  • Computer Vision and Semantic Search

    AI quality inspection, plus hybrid retrieval that combines vectors with keyword search.

  • Large-scale information retrieval

    Searching large document sets by meaning, not just by exact words.

  • AI analytics and recommendations

    Using generative AI to analyse business data and recommend next steps for users.

  • Digital Signature and quota management

    Electronic signing integrated with external systems, plus quota management with reporting.

  • Data Engineering

    Streaming and batch data architecture for enterprise analytics.

Executive team

  • CEO & CTO

    Sets technology direction, leads teams and governs enterprise systems, with experience in AI systems covering RAG, fine-tuning and semantic search, government and digital system architecture such as e-Licence, digital signature, NSW/ASW integration and e-payment, and data and infrastructure covering big data, data engineering and cyber security systems.

  • Lead Software Engineer & QA Lead

    Leads development teams, quality assurance and enterprise delivery, with experience in AI across RAG, semantic search and NLP, electronic services including e-Licence, e-contract, digital signature and NSW/ASW, web/mobile apps along with API gateway and OIDC/OAuth2/SSO, and data and data visualization.

Flagship product

Swarm Manager — a GPU and MIG management platform for the enterprise

See Swarm Manager in detail →
  • MIG partitioning across one card
  • Five-dimensional quota control
  • Automatic return of GPUs to the shared pool (Smart Hibernate)

Who we serve

  • 01

    Cloud & IDC Providers

    Turn AI servers into a multi-tenant rental pool, with capacity for resellers and metering of real usage.

  • 02

    Universities & Research Labs

    Five-dimensional quotas for researchers and students, with GPUs returned to the pool automatically when idle.

  • 03

    AI Startups & SMEs

    Rent GPUs or MIG slices sized for fine-tuning and inference, with ZFS storage and SSH.

  • 04

    Enterprise & Gov IT

    Private AI Cloud, RAG and AI software development on a complete security baseline.

Our company

GPU Plexus Co., Ltd. builds AI software, analytics systems and GPU infrastructure in Thailand with a team of Thai engineers.

About us →

Ready to design AI infrastructure that fits your organisation?

Tell our engineers about your challenge and we will help assess the right approach and cluster size.