PyTorch Meetup Singapore
We are excited to host the first PyTorch Meetup at the Red Hat Singapore office, bringing together customers, AI/ML professionals, engineers, researchers, and community builders for an evening of learning, collaboration, and open source exchange.
This meetup will focus on practical and emerging areas across the PyTorch ecosystem, including inference, distributed PyTorch, AI workloads, testing strategies, CI/CD pipelines, and how teams can build reliable, sc
The office is gated at the level 1 lobby so , we will typically send the building QR code to all the registered attendees which will bring them to level 46. We will send the qr code by noon of that day.
The meetup area is within our office, thus our global workplace policy requires them to sign in, so we will register them on our envoy system and they will also get an email to register themselves. The email will be sent usually done by noon. Luma should give us all the registered emails.
alable, production-ready ML systems. We will also highlight community initiatives and explore how contributors, developers, and organizations in Singapore can get more involved with PyTorch and the broader open source AI ecosystem.
Whether you are deploying models in production, optimizing inference, building distributed training workflows, improving testing infrastructure, or looking to grow an AI community, this meetup is designed to create meaningful conversations and hands-on connections.
Join us to learn, share, network, and help shape the future of the PyTorch community in Singapore.
Registration
Please complete your registration by filling out the following form:
https://forms.gle/Q2jfLNZDY5hS59RLA
Submission of the form is mandatory, as it is required for arranging your entry pass for the venue. Completing it in advance will help ensure a smooth and hassle-free entry experience at the event.
Event Lineup
Sudhir Dharanendraiah (Red Hat)
Sovereign Intelligence: Architecting APAC's AI Future
As APAC transitions from AI consumers to architects, Sovereign AI has become the critical foundation for regional technological independence. This talk dives into the engineering "plumbing"—including OpenReg, torch.compile, and FSDP—that scales high-throughput tools like vLLM on local hardware. Join us to learn how we are building a collaborative community to move open-source AI from research notebooks into a sovereign, production-ready reality.
Ziqi Zhao (Inferact)
Introducing vLLM's new Rust front-end
vLLM’s new Rust frontend: the motivation behind it, how it fits into the existing vLLM serving architecture, early integration and performance learnings, and the roadmap discussed in the RFC.
Pin Siang Tan (Embedded LLM)
Introduction to vLLM — High Performance LLM Serving & Project Update
vLLM is the most popular engine for high-performance LLM serving. In this talk, we’ll introduce the core architecture and features that make Large Language Model (LLM) inference and serving fast, easy, and cost-effective for everyone. Whether you are new to the engine or running production workloads, join us to explore the latest feature updates and the 2026 roadmap.
汪志鹏 (Wang Zhipeng) — Rakuten Asia Pte Ltd
vLLM-Omni: A Unified Platform for Omni-Modal LLM
vLLM-Omni evolved from vLLM into a standalone omni-modal serving framework with native support for autoregressive and diffusion models, unified multi-stage pipelines, and modality-aware scheduling. It now powers image, video, speech, omni, VLA, and world models behind a single OpenAI-compatible endpoint, while expanding toward embodied AI and multimodal RL with verl-omni.
Ayush Satyam (Red Hat)
torch.compile in the Wild
A source-code survey of how 10 popular PyTorch ecosystem projects—spanning inference, training, distributed, RL, and domain-specific frameworks—actually integrate torch.compile. Covers compilation boundaries, distributed wrapper ordering, graph break workarounds, and patterns you can apply to your own projects.
Sumantro Mukherjee (Red Hat)
PyTorch Community and CI
This talk covers the progress and updates about the PyTorch open-source community and the CI systems that keep the framework reliable across platforms and hardware. We will share insights into large-scale testing, contributor workflows, and the engineering challenges behind sustaining the PyTorch ecosystem.