The K Economy · Techvisory
Evidence base · industries and occupations

Industries and careers on the K

Named industries and named occupations placed on the evidence. The split is not between industries or between jobs. It runs inside them — by seniority, by task type, and in Malaysia's case straight through the diploma-holding middle.

Data-centre revenue
+43.1%
But Johor's 47 facilities are ~2,500–10,000 permanent jobs in total.
Computer services revenue
+3.7%
The boom is physical infrastructure, not software services.
Inside one Penang building
RM2,400
Equipment technician floor, against RM6,000–9,000 for a verification engineer. National median is RM3,027.
Skilled jobs created
−4.2%
Year on year, Q1 2026 — while total job ads rose 33%.
Read “ICT” carefully

Data centres +43.1% · computer services +3.7% · telecoms +3.9% · e-commerce income +2.9%. Treating “tech” as one sector is the single most common error in Malaysian workforce planning right now.

Industries, by arm

Upper arm

  • Data-centre activities — revenue +43.1%, but roughly 30–50 jobs per facility
  • E&E and semiconductor — IPI +26.4%; 48.2% of H1 2026 exports
  • Mining & quarrying — median RM9,678
  • Transportation & storage — +9.4% revenue, +4.3% employment; fastest-hiring
  • Construction — data-centre driven; M&E pay jumps of 20–50% between projects

Split — high pay, high exposure

  • Insurance & takaful — 91.8% exposed, highest in the country
  • Real estate — 87.4%
  • Finance & insurance — median RM7,700, 87.1% exposed
  • Recruitment & staffing — 78.6%
  • Information & communication — median RM6,000, but computer services only +3.7%
  • Global business services / shared services

Lower arm

  • Wholesale & retail — median RM2,600, below national
  • Food, beverage & accommodation — RM2,164, the lowest
  • Textiles & apparel — RM2,245
  • Agriculture — RM2,327
  • Administrative & support services

Occupations

Upper arm — monthly ranges, Malaysia 2026 recruiter guides

RoleRM / monthNote
Security Architect18,000–35,000Top of the published range in Malaysia
Cloud Architect15,000–25,000
Machine Learning Engineer15,000–23,000
Data Engineer12,000–20,000
Data Centre Design Manager12,000–15,000
AI Engineer10,000–18,000One of only three MyMahir emerging roles named publicly
DevOps / SRE8,000–18,000
Senior Process Engineer, advanced packaging7,000–12,000
IC Design Engineer6,000–15,000Hardest role to fill in the country
Equipment / Automation & Mechatronics Engineer4,500–8,000The realistic conversion target for a technician

Recruiter salary guides are estimates, not survey statistics. DOSM medians are official; Q1 2026 figures use a March reference month.

The hollowed middle

  • Software developer / application programmer — AI exposure index 1.0, the maximum, and on the Critical Occupations List
  • Financial analyst — exposure index 1.0
  • Computer systems analyst and web & multimedia developer — more than half their skills in the top exposure quartile
  • Technicians & associate professionals — 988,000 workers, 56% in the top quartile
  • Semiconductor equipment technician — RM2,400–3,500, below the national median
  • Manufacturing technician — RM2,500–3,500
  • Data centre technician — RM3,500–4,000

Lower arm — declining occupations

Bank teller
35%
Data entry clerk
34%
Cashier
30%
Admin assistant / exec secretary
28%
Accounting, bookkeeping, payroll clerk
24%
Claims adjuster
19%
Telemarketer
16%

Projected change to 2030, WEF Future of Jobs. Cashier is the largest absolute decline globally. Clerical support in Malaysia is 1.65m workers at 92% exposure.

Sheltered

Physicians, dentists, physiotherapists, nurses, judges, welders, crane and plant operators, electricians and skilled trades. Task exposure scores here run from 0.05 to 0.10 — and in several of these professions accountability is legally non-delegable, which is a stronger protection than any task score.

The percentage trap

Fastest-growing by percentage

WEF to 2030: Big Data Specialists +110% · FinTech Engineers +95% · AI/ML Specialists +85% · Software Developers +60% · Security Management +55% · Data Warehousing +50% · Autonomous & EV, UI/UX +45% · IoT, data analysts, environmental engineers, InfoSec analysts, DevOps, renewable energy engineers +40%.

High percentage. Low volume. High wage.

Largest absolute gains

The same WEF report: farmworkers +34m, delivery drivers, construction workers, shop salespersons, nurses, teachers.

High volume. Low wage. Only software developers appear on both lists — and software developers score maximum AI exposure.

Where Techvisory works

The same evidence, read as a targeting list. Ranked by the size of the exposed population against the capacity to fund a response.

#SegmentSize / exposureWhat we sellFunding route
1Insurance, takaful & financial services87–92% exposedRole redesign for claims, underwriting, compliance and back-office. AI governance for regulated data. Not tool training.Services uses only ~80% of levy against manufacturing's 100%
2Clerical & admin, all industries1.65m, 92% exposedThe levy-funded transition programme. Volume, standardised, sold to HR.Claimable; RM1,750/pax/day public ceiling
3Technicians & associate professionals — the M40988,000, 56% exposedM40 Defence: role redesign, human-edge skills, supervising AI. Nobody sells here.Claimable, plus 50% additional tax deduction on MSME AI training
4Global business services / shared servicesMyMahir Phase 1 sectorProcess automation assessment, workforce redesign, AI operating model.Large firms use ~96% of levy — compete on subject, not budget
5Semiconductor & E&E technicians13,679 against a 60,000 targetTechnician-to-engineer conversion: automation, PLC, robotics, IIoT.ETSI, PTPK, K-Youth, HRD Corp
6Data centre facilities, M&E and construction47 Johor facilitiesCertification pathways, project and planning management, DC operations. Premium, small cohorts.Employer-funded and claimable
7Recruitment, staffing & professional services78.6% exposedPartnership or white-label — they are disrupted while selling talent solutions, and they have the distribution.Channel play
8Retail, F&B, micro-enterprise~580,000 establishmentsLow-ticket, standardised, high-volume. Cost framing, never technology framing.PLM and the RM5,000 MSME digital grant. No premium product.
Sources and caveats

Sources. World Bank & ISIS Malaysia AI exposure brief (July 2025, LFS 2021) · DOSM Salaries & Wages and Employment Statistics · PERKESO EIS · MIDA · Knight Frank Data Centre Atlas 2026 · WEF Future of Jobs · Indeed Hiring Lab · Malaysian recruiter salary guides 2026 · TalentCorp / MyMahir · Critical Occupations List.

Caveats. The NY Fed disputes the AI reading — declines in AI-exposed postings began before ChatGPT and stabilised after 2023; fewer than 10% of workers have exposure above 0.4 and 40% have zero. Forecast is not measurement: WEF forecasts Big Data Specialists at +110% while observed Indeed postings put Data & Analytics at −39.6% against pre-pandemic, the worst of any category. Malaysia's three employment datasets are not interchangeable (LFS 16.82m employed; formal registered private-sector 9.03m filled jobs; wages survey 7.04m). MyMahir's 60 emerging roles are not public — only three are named anywhere: AI Engineer, Sustainability Specialist, Bio-process Engineer. No Malaysian declining-occupations list exists from job-posting data; PERKESO retrenchment-by-occupation is the proxy used throughout. Skilled-trades wage claims are unverified; the hard datapoint is US civil engineering postings at +154% against pre-pandemic.

Place yourself, then skill up

The map above is the prerequisite. The lead training turns it into a path; the AI courses at ai-courses.edisontkp.com are the skill layer.

Employers: occupation-by-occupation placement · opportunity map.