AI Work Index

Structural pressure

17%

Moderate Risk

Likely range

0–37%

Engineering Manager

Leads engineering teams, manages delivery, and drives technical strategy

This model suggests AI is more likely to enhance Engineering Manager than replace it. high exposure, but strong human bottlenecks mean AI augments rather than substitutes.

Compare

Risk depends on your actual work split

0%
37%
Augmented Medium Confidence 3-component blend
Higher risk than 59% of occupations
Pressure 17%
Market 77%
Confidence 52%
Human Moat High

Why This Score

AI Capability Overlap Increases risk

68% of tasks overlap with current AI

Human Coordination Reduces risk

67% human advantage from judgment & presence

Local Hiring Demand Reduces risk

77% demand buffer from SG labour market

Observed AI Adoption Reduces risk

AI usage 13pp below theoretical exposure

Official Demand Signals Reduces risk

Listed on SOL & JiD — active government demand recognition

These factors combine multiplicatively — larger bars do not mean proportionally larger contributions to the final score.

Net Displacement Risk 17%

Blended across 3 occupations using the same score logic as an occupation page. How this works

Tasks AI can handle

Code generation, test writing, documentation, code review suggestions, and debugging common patterns.

Where humans stay essential

System architecture decisions, complex debugging in production, cross-team coordination, requirements gathering, and security-critical code review.

Skills to focus on

System DesignDebugging Complex SystemsStakeholder CommunicationSecurity Awareness

Common tools in similar work

Adobe PhotoshopAdobe AcrobatAdobe Creative Cloud softwareApple macOS

Blended from O*NET matches across 3 component occupations.

Singapore Reality

Current Singapore signal

Labour now, industry footprint, and a directional 12-month read.

Weak market

The Professionals, Managers, Executives & Technicians labour market is weak. Vacancy rate is 3.1% and was essentially flat versus last quarter. recruitment is running above resignation (1.5% vs 0.9%). live job ads are active, with 26 visible postings in the last 30 days. employer pressure is moderate.

Vacancy rate

3.1%

↓ 3.1% year-on-year

Hiring balance

1.5%

recruit vs 0.9% resign

Retrenchment

1.5 per 1,000

Low incidence

Professionals, Managers, Executives & Technicians data · Q4 2025 full

Top Industries

Where this work is concentrated

Financial & Insurance Services

↓ cooling
15%

Public Administration & Education Services

14%

Professional Services

→ stable
12%

Top 4 vacancy sector

Industry vacancy overlays use the latest published detailed cross-tab (2025 QQ), which can lag the main labour monitor.

12-Month Outlook

Rule-based, not a prediction

Stable
Displacement Resilient
Augmentation Under Pressure
Demand Resilient
Wage Pressure Resilient

What To Do Next

Offset potential: Medium Separate from the core score. This is a blended estimate of how much demand, redesign room, and transition support could cushion pressure.

This estimated role shows some offset potential, but it depends on demand and transition pathways holding up across the blended occupation set.

Demand support: Medium Transition support: Medium Reallocation room: High Switching friction: High

What helps

  • A meaningful share of the work can likely be reorganized around AI rather than removed outright.
  • Demand still persists through current labour or hiring signals.

What could slow it down

  • Current demand support is thin, so offsets may take longer to show up.

Blended from 3 component occupations. This remains a heuristic support layer rather than a direct measure of realised reinstatement.

Blended from 3 component occupations; anchored on Software and applications manager.

Compare with similar roles or occupations

Compare with... →

Structural AI exposure scores, not employment predictions. Methodology

V4.1 structural score · 2024 wages · Q4 2025 full live monitor · 562 occupations · 88 roles

Latest official labour release: MOM Labour Market Report Q4 2025 (2026-03-20)

Made by Kirill So · Built with Claude & GPT
MIT Licensed · Open Source · Structural release V4.1 · Updated 2026-03-20