Our company
A small team that takes
data work seriously
We're a Kuala Lumpur-based applied AI practice — working with businesses that want practical, well-reasoned approaches to their data challenges.
← Back to HomeOur story
Where Aethernet comes from
Aethernet began in 2019 when two engineers — one from the fintech sector, one from academic machine learning research — noticed the same recurring problem: organisations had data but lacked the internal capacity to make structured sense of it.
Rather than build a large consultancy, we kept the practice focused. We take on a limited number of engagements at a time, which lets us give each project the attention it deserves. Every model we build, we also explain — documentation and knowledge transfer are part of what we deliver, not afterthoughts.
Based in Kuala Lumpur, we serve clients across Malaysia and occasionally work with regional teams in Singapore and Indonesia. Our work spans financial services, e-commerce, healthcare operations, and professional services.
Our mission
What we set out to do
Make applied AI work accessible to businesses that aren't ready to hire a data team but have real problems where AI can help. We aim for clarity: in the work itself, in how we explain what we're doing, and in how we price our services.
Honesty about scope
We'll tell you when a problem doesn't actually need AI, and what might serve you better.
Methodical, documented work
Every project ends with clear documentation so your team understands what was built and why.
Sustained relationships
Many of our clients return for follow-on work. That continuity shapes how we approach every first engagement.
The people
Who does the work
Reza Harith
Co-Founder & ML Lead
Former ML researcher with a background in predictive modelling for financial services. Leads model design and performance evaluation on every engagement.
Syafiqah Nor
Co-Founder & Data Engineering
Specialises in data pipeline architecture and systems integration. Ensures that what we build connects cleanly to what clients already have.
Amir Mokhtar
Senior Analyst
Works closely with clients during the scoping and data-review phase. Translates business questions into structured problem statements for model development.
How we work
Standards we hold ourselves to
PDPA Compliance
We operate in accordance with Malaysia's Personal Data Protection Act 2010. Client data is used only for the agreed scope and not retained beyond project completion.
Documented Deliverables
Every model and system comes with written documentation covering architecture, data flows, and maintenance guidance. No black-box handovers.
Baseline Comparisons
We always compare our models against simple baselines. If the improvement doesn't justify complexity, we say so before proceeding.
Version-Controlled Work
All code is version-controlled and handed over with your project. You own what we build and can extend it independently or with another team.
Scope Agreements
We write clear scope documents before starting. Changes to scope are discussed openly and any pricing implications are communicated before work proceeds.
Knowledge Transfer
Towards the end of each engagement, we walk your team through what was built and how to manage it going forward.
Applied AI in Malaysia
What we believe about this kind of work
AI adoption in Malaysian businesses is at a meaningful point. Organisations have more data than ever, and the methods for processing that data have become substantially more accessible. The barrier now is less about technical availability and more about knowing which problems are actually solvable, what data quality is needed, and how to embed outputs into day-to-day operations.
We think the work of applied AI is mostly careful, undramatic problem scoping — deciding what to build, what data to use, and what would constitute a meaningful result. The modelling itself tends to follow naturally from that groundwork. Teams that struggle with AI projects often do so because the problem definition was vague, not because the models were technically insufficient.
That orientation shapes how Aethernet engages with clients. We spend time at the beginning making sure the problem is clear and the data is adequate before any model development begins. It slows the start, but it improves outcomes substantially.
Work with us
Straightforward about what's possible
If you have a data problem and you'd like an honest assessment of where AI fits, get in touch.
Contact Aethernet