Installed is not working
The machine in the people function. What AI has actually been deployed across HR in mid-market and enterprise, globally and in India, what is holding weight, and the five gates a use case has to clear before it is worth deploying at all.
Installed is not working
The machine in the people function. What AI has actually been deployed across HR in mid-market and enterprise, globally and in India, what is holding weight, and the five gates a use case has to clear before it is worth deploying at all.
The evidence behind the answer
What it covers
AI deployment across the people function in mid-market and enterprise organizations, globally and with a separate read on India. Four deployment architectures, what each one costs to run, where each one stalls, and the split between what mid-market organizations are doing and what enterprises are doing. It covers hiring, service delivery, payroll, learning, performance, planning and employee relations.
How it was built
Public disclosure read closely, vendor material taken apart, deployment claims traced back to the organizations named in them, and what we see directly in our own workshops and advisory work. Thirty-four sources are cited in full. Where we are inferring rather than observing, the report says so on the page rather than in a footnote.
The instruments inside it
Three frameworks were built for this report and are published in it rather than held back. The Reality Curve, which replaces the hype cycle by plotting realized value against depth of system dependency. The Load Map, a two-by-two of evidence against plumbing depth. And the five readiness gates, which is the part most readers photograph.
What it is not
Not a vendor comparison, not a buyer's guide, and not a ranking. Nobody paid to appear in it and no vendor saw it before publication. If you want to know whether a specific product has a confirmed deployment, that is the register's job and this report will point you there.
Depending on which chair you are in
You will be asked what the AI spend bought. The report gives you the honest shape of what has worked across the function, so the answer is a position rather than a scramble.
Your suite shipped AI you did not choose. The report separates what those modules actually carry from what they are licensed to do, and gives you the five questions to put to your vendor before the next renewal.
HR is where your governance model meets language. The report sets out where that breaks, what the jurisdictions do to it, and which deployments cleared it.
Category heat and category evidence point different ways. The report shows the spread across eight categories with the counts attached.
The structure, in order
Find the section you need
How it was built
Public disclosure read closely, vendor material taken apart, deployment claims traced back to the organizations named in them, and what we see directly in our own workshops and advisory work. Thirty-four sources are cited in full. Where we are inferring rather than observing, the report says so on the page rather than in a footnote.
Download the report
Free to use in your organization
The file is delivered here as soon as it is available. Your details are never shared with a vendor.
The file has not been uploaded yet.
Also worth your time
Identity and onboarding agents in retail banking
Nineteen claimed deployments checked with the banks named in them. Seven confirmed at full production, four at pilot, and eight that did not survive the second question.
The Voice Agent Map
Sixty-one companies selling enterprise voice agents, each carrying a durability verdict rather than a quadrant position, with the evidence behind the verdict.