Where the work lands — five-workforce distribution
Task inventory
Process
Task
Role
Target workforce
Automation potential
Skills shift
Skills shift — what to keep, build and retire
Keep Judgement & relationships
Hiring-manager consultation and intake framing
Right-to-work & eligibility verification
Final payroll sign-off and accountability
Sensitive employee conversations
Exception handling the model can't resolve
Build New capabilities
Prompting & reviewing generative drafts (vacancy, contract text)
Supervising AI agents across systems (orchestration oversight)
Exception triage & data-quality stewardship
Reading automation dashboards and control signals
Designing and owning the controls that automation runs on
Retire Fading manual work
Manual re-keying between systems
Copy-pasting and saving files to folders
Sending standard, templated notifications by hand
Manual error-list logging and status chasing
Assembling documents from data by hand
DIRECTIONAL
Method. Each task from the four process maps is reallocated to its best-fit workforce and given an automation-potential score — the estimated share of the task's effort that shifts off a human with today's technology. Figures are directional estimates for discussion, not a headcount business case.
Technical automation potential is not a headcount saving: after controls, exceptions and adoption, realistic near-term effort freed is lower (a 35–50% range) and is best reinvested to absorb growth. Hardening these figures into € and FTE needs FTE by role/location, task volumes and cycle times, and the current tech stack. All data on this page is synthetic and illustrative.