About the firm
A firm that reads
before it speaks
Ondoria was established to offer something less common than speed: careful attention to the particular situation of each organisation that engages it.
← Return Home§ I — The Firm
How Ondoria came to be
Ondoria was founded in Kuala Lumpur by practitioners who had spent several years observing how organisations in Malaysia were approaching applied AI. What they found, again and again, was that the difficulty was not technical capacity — it was the absence of clear thinking about what the organisation actually needed and what constraints would shape any workable system.
The name Ondoria is drawn from an older word for a place of careful reflection. It was chosen deliberately: the firm's position is that considered framing of a problem is itself a form of work, and that organisations benefit from having a steady, independent reader alongside them as they develop their own applied capability.
The firm takes on a small number of engagements at any one time. This is not incidental — it reflects a view that counsel of the kind Ondoria offers is only useful when it is genuinely attentive. Reading a situation well requires time, and time requires limits on volume.
Ondoria holds no equity in client systems, takes no operational role, and does not build proprietary tools for resale. Its orientation is toward the client's capability — not toward the firm's own accumulation.
§ II — Orientation
What guides the work
Attention over volume
The firm limits the number of engagements it accepts. Each client situation receives the full attention it warrants, not a portion of a divided calendar.
Written as the primary medium
Verbal exchange is useful but transient. Ondoria works in writing because documents persist — they can be returned to, revised, and shared within the client organisation.
No operational entanglement
The counsel reads, reflects, and returns an opinion. It does not execute, does not manage systems, and does not create dependency on its continued presence.
Capability stays with the client
Every engagement is designed so that what is learned — the framing, the reasoning, the record of the pilot — remains within the client organisation after the engagement closes.
§ III — People
The people behind the work
Ahmad Mazlan
Principal Counsel
Ahmad leads client engagements, with a background in operational research and applied statistical work across the financial and logistics sectors in Malaysia.
Nurul Rashidah
Research Associate
Nurul supports the written work — drafting papers, preparing schematics, and maintaining the analytical record across engagements. She holds a postgraduate qualification in computational linguistics.
Teoh Kai Wen
Technical Associate
Kai Wen conducts the technical work in Quiet Pilot engagements, working alongside client engineering teams. His experience spans model evaluation, data pipeline design, and system documentation.
§ IV — Standards
How we hold ourselves to account
Confidentiality
All client materials are treated as confidential. We do not discuss client situations externally and operate under formal confidentiality agreements for all engagements.
Written record
Every engagement produces a written record. No work concludes without a document the client can keep — whether a reading paper, a pilot record, or a quarterly reflection.
Data protection
We handle personal data in accordance with Malaysia's Personal Data Protection Act 2010. Client data is not retained beyond the engagement period without explicit agreement.
Responsible use
We decline engagements where the application of AI models would, in our assessment, create undue risk to individuals or communities. This position is stated plainly in our terms.
Continuing development
Associates maintain active engagement with published research in applied machine learning, natural language processing, and the governance of AI systems in commercial settings.
Client feedback
At the close of each engagement, we invite a written response from the client. This is kept on file and informs how we approach future work — it is not used for marketing purposes.
§ V — Expertise
Applied AI work in the Malaysian context
The application of machine learning and language models within Malaysian organisations presents a specific set of considerations — multilingual operating environments, a range of data maturity across industries, and a regulatory landscape that is still developing its position on AI governance. Ondoria works within this context, not in spite of it.
Our counsel draws on experience across financial services, manufacturing operations, professional services, and the public sector. In each case, the questions that matter most are not purely technical: they concern what the organisation is trying to achieve, what data it actually has, and what decisions a model's outputs would feed into.
The written engagement format — the Applied Reading — emerged from practice rather than theory. It reflects the observation that most organisations benefit from a clear statement of their situation before they commit to a technical path. The reading does not advocate for any particular approach; it describes what is present, what is absent, and what questions would need to be answered before a sound choice could be made.
Ondoria holds no partnership with model providers, cloud platforms, or software vendors. This independence is a deliberate feature of the firm's position: our assessments are not shaped by commercial relationships with technology suppliers.
§ VI
Write to us with your situation
There is no obligation attached to writing. We read every message and respond in kind, usually within two working days.
Begin a Conversation