the queue between a question and its answer
self-service analytics never removed the wait for workforce answers, because the slow part was never retrieval.
blog
ideas on agentic AI
ideas on agentic AI
and the future of HR
perspectives, research, and playbooks from the team building agentic intelligence for performance, skills, and growth.
enzo
performance
goal quality is measurable, and most goals fail the test
a goal nobody can review produces no feedback, and that failure can be scored, sampled and tracked across an entire organisation.
engagement
your engagement survey measures the wrong latency
engagement moves day to day, the annual survey reports once a year, and the gap between those two clocks is where most listening programmes fail.
wellbeing
burnout is a job design problem, not a resilience problem
the evidence points at demands, resources and control, yet most organisations respond by measuring how well people cope with them.
analytics
the workforce report your board and your auditor both accept
ISO 30414 became a requirements standard in 2025 and the EU narrowed CSRD in 2026, which changes what a credible disclosure set looks like.
skills
career paths people can walk
career frameworks name destinations and skip the route; paths built from skill distance tell people where they can go next.
workforce planning
workforce planning is a decision, not a spreadsheet
build, buy, borrow or bot: every capability needs a supply decision on the record, with the evidence that would reverse it.
frameworks
why competency frameworks die in month seven
role requirements drift faster than any rewrite cycle; a framework survives only if it is maintained monthly like an operating system.
assessment
assessing skills at scale without fooling yourself
validity is an argument you build, reliability is the ceiling you cannot exceed, and most internal assessments fail on both.
internal mobility
internal mobility runs on two clocks
one clock fills the role open this month, the other builds the career that takes three years, and most organisations wind only one of them.
succession
succession slates need evidence, not ratings
nine-box grids produce agreement quickly and accuracy rarely, and a board question will find the difference every time.
learning
learning spend needs a chain of evidence
training evaluation stops at satisfaction scores because the next four links are unowned, not because they are impossible.
axon
your skills taxonomy is a dictionary, not a diagnosis
a taxonomy resolves what to call things, and cannot tell you proficiency, context or currency, which is where every workforce decision lives.
enzo
autonomy should be a dial, not a switch
a four-level model for governing AI in HR, with what to log at each level and the evidence required to move a use case up.