SYS: DAMAI-LOGIC-01 / BENEFITS DAMAI LOGIC +60 3-9145 8736
DAMAILOGQR.
Damai Logic working principles
// WHY DAMAI LOGIC

Working Principles That Hold Across Every Engagement

The practical advantages of choosing Damai Logic come from a consistent approach to scope, evaluation, and documentation — not from claims about algorithm sophistication.

Back to Home
// CORE ADVANTAGES

Six Things That Shape Every Engagement

Scope in Writing, Before Work Starts

You know what you're getting before we begin. If the scope changes, we agree to that change explicitly. There are no surprises about what was and wasn't included.

Evaluation Against a Stated Baseline

Models are compared to a simple baseline agreed before modelling begins. If the model doesn't improve on it, we say so. You're not sold a model that can't justify its own existence.

Documentation That Enables Handover

Build deliverables include technical documentation covering the data pipeline, retraining process, and failure modes. Your team can operate the system without ongoing dependence on us.

Uncertainty Shown, Not Hidden

Forecasts include uncertainty intervals. Anomaly detectors have a documented false-positive rate. We don't present outputs with more precision than the underlying data supports.

Complexity Proportionate to Need

We use the simplest model that meets the requirement. More complex approaches are introduced only when simpler ones demonstrably fall short on held-out data.

Education That Transfers Practical Judgement

The Reading Group is led by practitioners. Sessions are grounded in current methods in production use. Participants leave with judgement they can apply independently, not just familiarity with terminology.

// DETAILED BENEFITS
EXPERTISE

Practitioners, Not Generalists

Each member of the Damai Logic team has worked on operational data problems in a specific domain before joining the firm. Forecasting engagements are led by someone who has built forecasting systems for manufacturing and retail operations. Detection engagements are led by someone with direct experience in infrastructure monitoring and transaction surveillance.

  • Domain-matched experience for each engagement type
  • MSc and above educational backgrounds
  • Active familiarity with current methods, not dated curricula
METHODS & TOOLS

Current Methods, Chosen for Fit

We work across statistical time-series methods, gradient-boosted approaches, and neural sequence models, selecting based on the characteristics of the client's data. We don't have a preferred algorithm that we apply regardless of the problem — the method follows from the data and the requirement.

  • Statistical and ML approaches evaluated on the specific dataset
  • No vendor lock-in — deliverables use open-source tooling where possible
  • Infrastructure requirements scoped to what the client already has
CLIENT ENGAGEMENT

Clear Communication Throughout

We maintain a weekly written update through every build engagement. If we encounter a problem that will affect the timeline or scope, we raise it at the next update — not after the deadline has passed. We also tell clients early if a project is unlikely to produce a useful result.

  • Weekly written progress updates
  • Issues raised proactively, not retrospectively
  • Honest feasibility assessment at the start
PRICING & VALUE

Fixed Fees, No Hidden Variables

Engagement fees are stated upfront and do not change unless the scope changes. There are no success fees, hourly overruns, or licensing charges embedded in deliverables. Cloud compute costs, where applicable, are estimated before work begins and agreed separately.

  • Forecasting System Build: MYR 2,090 fixed
  • Anomaly Detection: MYR 1,360 fixed
  • Reading Group: MYR 480 fixed
RESULTS

Measured Against What Matters

The success criteria for every engagement are stated before work begins — agreed with the client and written into the scope document. For forecasting, this is typically a forecast accuracy metric on held-out data, compared to the baseline. For anomaly detection, it is the false-positive rate and the detection rate on historical incidents.

  • Success criteria agreed before the engagement begins
  • Results reported against pre-agreed criteria
  • No post-hoc reframing of what success means
// COMPARISON

How We Compare to Typical ML Providers

Feature Typical Providers Damai Logic
Scope defined in writing before start
Evaluated against a stated baseline
Uncertainty intervals included in forecasts
False-positive rate documented for detection systems
Deliverable your team can operate independently
Fixed fee, no success-fee or hourly overrun
Education led by practitioners with production experience
// WHAT SETS US APART

Distinctive Features of Our Approach

The Limitations Document

Every build deliverable includes a section describing the conditions under which the system's outputs should not be trusted. This document is not boilerplate — it is specific to the system and the data it was trained on. Most providers don't include this; we consider it a core part of the work.

The Baseline Veto

If a model we build doesn't improve on the agreed baseline on held-out data, we report that clearly and do not invoice for a working system we haven't produced. The baseline comparison is not a formality — it determines whether the engagement has achieved its technical objective.

Reading Group as Structured Inquiry

The Reading Group is not a survey course. Each session is built around selected readings and discussion questions developed by our team. Participants are expected to read before arriving and to contribute — the format assumes competent technical people, not passive recipients of information.

Operations-Calibrated Scope

Our engagements are sized to what an operations or small data team can realistically maintain. We don't build infrastructure that requires a dedicated ML team to keep running. The output of every build engagement fits into the client's existing technical environment.

// MILESTONES

Firm Milestones

6
Years Operating
40+
Engagements Completed
12
Reading Groups Delivered
MY
Peninsular Malaysia

MDEC Registered ICT Firm

Registered with Malaysia Digital Economy Corporation as an ICT service provider.

MITI SME Technology Partner

Recognised under the Ministry of International Trade and Industry's SME technology partner programme.

Industry–Academia Collaboration

Active collaboration with the Faculty of Computer Science and Information Technology, Universiti Malaya.

// NEXT STEP

See These Principles in Practice

The best way to understand how we work is to have a preliminary conversation about a specific problem your business is facing. There's no charge for that conversation.