Government desk work

Advisory advice that stays inside government boundaries.

Our engagements are built for the public sector's actual working conditions — not adapted from commercial deployments. That distinction matters when the files involve citizens.

Back to Home

What You Gain from Working with Pelita

Public Sector Familiarity

Our advisers have worked inside Malaysian government modernisation programmes. We understand how cases are received, routed and resolved — including the informal process steps that no vendor diagram captures.

Compliance Built In

PDPA obligations and MAMPU guidance are part of how we design every engagement — not a checklist added at the end. Your usage policy and data handling approach are written for your specific agency context.

Officer Control Maintained

No AI output in our deployments can trigger an action without officer approval. Every correspondence draft and case summary is reviewed by a named officer before leaving the system.

Plain Written Briefs

Every engagement produces a written brief your director can read without specialist knowledge. We do not deliver dashboards, proprietary portals or slide decks that require a vendor to interpret.

Air-Gapped Capability

For agencies handling restricted or sensitive classifications, we can design an integration that operates entirely within your existing government infrastructure — no data leaves approved systems.

Paced to Your Cycle

Government procurement, endorsement and approval cycles are not obstacles — they are the schedule. We build our engagement timeline around yours, not around a commercial quarter-end.

Genuine Public Administration Background

The advisory team at Pelita holds direct experience from government-facing projects under Malaysia's MyDigital and MAMPU capacity-building programmes. We have sat in the same briefings your operations heads attend, and we have read the same MAMPU circulars that govern AI deployment in public agencies.

This means we do not need to translate between government practice and technology theory. We work within the vocabulary your agency uses — official correspondence formats, authorisation hierarchies, data classification procedures.

  • Familiarity with JPA, MAMPU and JKDM correspondence practices
  • Experience working with statutory bodies and state agencies
  • Understanding of Bahasa Malaysia and English official correspondence norms
  • Practical knowledge of GLC governance and board reporting requirements

Careful Selection, Not Maximum Deployment

We do not push a particular AI product. Our first task in a Service Reading Review is to assess whether the agency's situation calls for AI assistance at all — and if so, which category of assistance is appropriate for the correspondence load and data sensitivity involved.

Where deployment does proceed, we select tools on the basis of their suitability for government data environments, not on the basis of commercial partnerships. Air-gapped and read-only architectures are a standard part of our technical toolkit.

  • Vendor-neutral selection based on data environment requirements
  • Air-gapped and read-only integration architecture available
  • No data processing outside approved government infrastructure
  • Technical documentation written for internal IT teams, not vendors

One Adviser Throughout, Not a Rotational Team

Each Pelita engagement is led by the same adviser from the initial briefing through to the final written deliverable. The operations head and service-counter lead you meet in week one are still the people you call with questions in week six.

For Quarterly Stewardship clients, continuity across quarters means the adviser reading your audit outputs already understands the context from previous periods — so the brief your director receives is cumulative, not a standing update from a different team member each time.

  • Named adviser throughout the engagement
  • Continuity across quarterly stewardship periods
  • Direct contact with the adviser, not a support queue
  • Briefings scheduled around agency calendar, not our availability window

Pelita versus Typical AI Vendors

Many agencies receive proposals from commercial AI vendors who have adapted their products from private-sector use cases. The table below reflects what that typically means in practice.

Consideration Typical AI Vendor Pelita
Public sector background Adapted from commercial case studies Direct government programme experience
MAMPU alignment Mentioned in proposal, rarely designed in Integrated from first engagement day
PDPA compliance approach Standard data processing agreement Agency-specific data classification review
Officer authority May allow AI to trigger workflows Officer approval required at every step
Deliverable format Proprietary dashboard or portal Written brief, no login required
Data storage Vendor cloud by default Agency infrastructure; air-gapped available
Engagement pace Commercial quarter-end driven Follows agency procurement and approval cycle

Distinctive Features of Pelita's Work

A Brief, Not a Dashboard

Every engagement ends with a written document your director can act on, file with internal audit, or present to a committee — without needing to log into a vendor system or pay for ongoing access.

We Read the Actual Files

The Service Reading Review involves Pelita advisers reading a real sample of your agency's citizen correspondence — with appropriate authorisation — not modelling a notional workflow from a description.

Honest Assessment, Including "Not Yet"

If our review suggests the agency is not ready for AI assistance, or that the data environment makes it inadvisable, we say so in the brief. We do not shape findings to justify a larger deployment.

Quarterly Audit as Standard Practice

For ongoing stewardship clients, we audit AI outputs and boundary conditions every quarter and write a brief for the director — so performance and compliance are reviewed regularly, not assumed to hold.

Professional Recognitions & Figures

18+

Government and GLC engagements completed

6

State agencies with active Quarterly Stewardship

100%

Engagements completed with officer authority preserved

PDPA

All deployments aligned with Malaysian data protection law

MyDigital Advisory Network — Associate Member, 2024

MAMPU Public Sector AI Working Group — Contributing Adviser, 2024–2025

Malaysian Institute of Management — Affiliated Practice, since 2023

A Service Reading Review is a low-commitment starting point.

It tells you whether AI assistance belongs in your agency's workflow — and, if so, where. The brief we produce is yours to keep and act on as you see fit.

Request a Review