Quantitative Methods Applied
to Real Operational Questions
Damai Logic was formed to do a specific kind of work — building systems that let businesses extract usable information from data they already collect, without overstating what those systems can tell you.
Back to HomeHow Damai Logic Came Together
The firm started from a pattern we noticed across multiple consulting engagements in Malaysia: businesses that collected operational data — inventory movements, maintenance logs, transaction records — but lacked a structured way to extract forward-looking information from it. The gap wasn't access to sophisticated algorithms. It was the absence of someone willing to work through the data rigorously and be honest about what it could and couldn't tell you.
We opened in Kuala Lumpur because the city's mix of manufacturing, financial services, and logistics companies gave us a practical spread of the kinds of forecasting and detection problems we wanted to work on. We kept the firm small deliberately. A small team can maintain consistent standards; it can also tell a prospective client when a project isn't a good fit for quantitative methods.
The name Damai — meaning calm or peaceful in Malay — reflects something we try to bring to engagements that can otherwise generate anxiety: data projects frequently involve confronting the limits of what can be known. We think that confrontation is productive, and we'd rather have it early.
- Kuala Lumpur, Wilayah Persekutuan, Malaysia
- Founded 2019
- Applied Machine Learning & Data Science
- Focused on operating businesses
- English and Bahasa Malaysia
To give Malaysian businesses practical access to quantitative forecasting and monitoring without the speculative framing that often surrounds machine learning services.
The People Behind the Work
Ahmad Nazrin
Principal — Forecasting
Ahmad leads forecasting engagements and has worked on demand and capacity problems across retail and manufacturing clients in Peninsular Malaysia. His background is in applied statistics and operations research.
Suraya Tan
Principal — Detection Systems
Suraya builds and tunes anomaly detection systems, with particular experience in infrastructure telemetry and transaction monitoring. She holds an MSc in Computer Science from Universiti Malaya.
Rajan Krishnaswamy
Lead — Technical Education
Rajan leads the Reading Group programme and developed the curriculum from his experience as an engineering manager in financial services. He focuses on making current ML methods legible to technical practitioners without a specialised ML background.
How We Conduct Our Work
Written Scope Before Work Begins
Every engagement starts with a written description of the deliverable, the evaluation criteria, and the data requirements. If the scope changes during the work, both parties agree to the revision before it is acted on.
Baseline Comparisons
Models are evaluated against a stated baseline chosen before modelling begins. The baseline is selected jointly with the client. Results are presented in comparison to that baseline, not in isolation.
Data Confidentiality
Client data is not used outside the scope of the engagement. We sign data-handling agreements at the start of each project. Access is limited to team members directly assigned to the work.
Documented Deliverables
Build engagements include technical documentation covering the data pipeline, the model retraining procedure, known limitations, and the monitoring approach. Deliverables are designed to be handed over, not maintained indefinitely by us.
Stated Limitations
We include a limitations section in every deliverable. This describes the conditions under which the system's outputs should not be trusted, and what events might require a retraining or reassessment.
Practitioner-Led Education
Reading Group sessions are led by practitioners who have worked on the problems being discussed, not trainers presenting vendor materials. Discussion is the primary format; slides are secondary.
Quantitative Work for Malaysian Businesses
Damai Logic works with businesses that have a clear operational question and the data to begin addressing it. Our engagements are not advisory — we build systems and deliver education. The systems we build are evaluated against pre-agreed criteria and documented thoroughly enough for client teams to operate them independently.
Forecasting work at Damai Logic begins with data that already exists in a client's systems: sales records, production logs, financial statements. We assess that data for the kind of signal it contains, build models calibrated to the scale of that signal, and deliver outputs that include uncertainty ranges rather than false precision. For most operational forecasts, communicating the uncertainty is as important as the point estimate.
Anomaly detection engagements follow a similar pattern. We work with the operations team to define, in advance, what kinds of readings would warrant attention. That definition shapes the tuning of the detection system. A system tuned without that conversation tends to generate either too many alerts or too few; both outcomes reduce the value of the system.
The Reading Group programme responds to a specific gap we observed: senior technical staff at Malaysian companies who needed to evaluate vendor claims and architectural proposals involving machine learning, but who had not had the opportunity to develop familiarity with current methods. Six structured sessions, led by practitioners and grounded in selected readings, have consistently provided a workable foundation for that kind of evaluation work.
We operate from Kuala Lumpur and take on clients across Peninsular Malaysia. We bring direct experience in manufacturing, logistics, financial services, and telecommunications applications of quantitative methods.
Ready to Discuss an Engagement?
If you have an operational question and the data to begin working on it, we'd be glad to talk through whether a quantitative approach makes sense and what it would involve.
Contact Damai Logic