Building a Place Where Hardware Knowledge Flows Freely
sablefins started with a straightforward observation: people interested in AI accelerator technology had plenty of content to read but few places to think through it together. We wanted to change that.
How sablefins Came to Be
sablefins was founded in Kuala Lumpur by a small group of practitioners who kept bumping into the same problem: the public conversation about AI was moving fast, but the underlying hardware — the chips, the memory systems, the infrastructure that makes it all run — was rarely discussed with any depth in accessible, community settings.
We built sablefins as a kind of town square for that conversation. Not a course platform, not a conference — something more like a reading room with good company. A place where a first-year engineering student and a senior infrastructure professional can sit at the same table, share what they're working through, and learn from each other.
The name reflects that spirit. A sablefins is a quietly distinctive fish — not flashy, not loud, but worth knowing if you're paying attention. We thought that fit a community focused on depth over noise.
We've kept the community deliberately manageable in size. That's a choice. Larger platforms optimise for reach; we optimise for the quality of what happens when people sit down and work through a hard topic together.
Our Mission
To make AI hardware knowledge more accessible and more social — by giving curious people across Malaysia a shared space to read, discuss, and grow their understanding alongside others.
Our Approach
We focus on community over content volume. Discussion areas are organised and moderated. Reading paths are curated rather than crowdsourced. Peer circles are kept small so everyone has a voice. We believe slower, more thoughtful learning produces better outcomes than a firehose of material.
Our Values
- Openness — every question is worth asking
- Balance — depth over breadth, always
- Respect — for where each member is in their learning
- Honesty — we say what we don't know as readily as what we do
Who Keeps sablefins Running
A small core team looks after the platform, curates resources, and facilitates the peer circles. Everyone here is also a participating member of the community.
Ahmad Hafizi
Community Director
Has spent the better part of a decade following developments in compute infrastructure. Believes that the best way to understand a subject is to explain it to someone else.
Nurul Liyana
Learning Experience Lead
Designs the reading paths and structures the peer circle format. Background in curriculum development; joined sablefins because she wanted the community she wished had existed when she was first learning.
Rajan Subramaniam
Platform & Operations
Keeps the technical side running and manages community onboarding. Also facilitates one of the active Mentor Circles, covering chip architecture fundamentals.
How We Keep the Community Worthwhile
We hold ourselves to a set of practical standards that shape everything from how we curate resources to how we handle member data.
Curated Resources Only
Everything on the shared shelf has been reviewed by at least one team member before it's added. We'd rather have fewer, better resources than a large unchecked pile.
Facilitated Discussions
Discussion areas are actively moderated to keep conversations on-topic and respectful. Members can flag anything that doesn't meet community standards and expect a response within one working day.
Data Privacy
We collect only what's needed to run the community. Member data is stored on Malaysian-based servers and is never shared with third parties for marketing. We comply with PDPA requirements.
Small Circle Sizes
Mentor Circles are capped at eight members. This isn't just a selling point — it's a quality commitment. We will not open a circle until we have the right facilitator for it.
Transparent Membership Terms
Membership terms, pricing, and what's included are clearly stated before any payment is made. No hidden charges or automatic renewals without prior notice.
Member Feedback Channels
We run a quarterly feedback cycle where members can share what's working and what isn't. Changes to the platform and reading paths are informed directly by those conversations.
AI Hardware Literacy in Malaysia — A Growing Conversation
The computing infrastructure behind modern AI systems — GPU clusters, custom silicon, high-bandwidth memory, distributed training setups — is complex and fast-moving. For most practitioners in Malaysia, genuine understanding of this layer requires both sustained effort and access to people who have thought about it carefully.
sablefins fills a gap in the local ecosystem. While technical documentation and online courses provide raw material, the peer dimension — having somewhere to sit down with others working through the same questions — is harder to find. Our community is structured around that peer dimension as its core.
Members range from university students exploring hardware fundamentals to professionals in cloud infrastructure, data engineering, and hardware procurement who want a more rounded picture of the field. The common thread is genuine interest and a preference for depth over surface-level summaries.
Want to be part of the conversation?
Reach out and we'll walk you through the membership options. No pressure — just an honest conversation about whether sablefins is the right fit for what you're looking for.
Get in Touch