From HR to People Intelligence: Building the Future of Work with AI

People Intelligence is my term for the intersection of people data, technology, organizational context and human judgment. It is not another HR software category, a dashboard vendor's marketing label, or a rebranded people-analytics team. It is a way of running the People function so that every important decision about talent is informed by evidence, accelerated by technology, and still owned by a human being who understands what is at stake.

The old model asked HR to administer processes. The people-intelligence model asks HR to produce insight: Where do our best hires actually come from? Which managers are retaining talent and which are quietly losing it? What is the real cost of a slow promotion decision? These are business questions, not administrative ones. And they require the same rigor we expect from finance, product or engineering.

Traditional HR was built for a stable, co-located, single-country company. Its core responsibilities were policies and administration, recruitment, compliance, payroll coordination and employee relations. Success meant smooth processes: offers signed on time, files in order, grievances handled, audits passed.

That model wasn't wrong — it was built for its era. But it positioned HR as a support function that reacted to the business rather than a function that informed it. HR knew a great deal about people and very little, in structured form, about what made them join, stay, perform or leave. The data existed. It just lived in filing cabinets, inboxes and institutional memory.

Five shifts, compounding on each other, broke the old model open:.

• Technology — cloud HRIS, ATS and collaboration tools turned every people process into a stream of data.
• Remote work — distributed teams made 'management by walking around' impossible and forced intentionality about how work and culture actually happen.
• Global teams — talent stopped being local. A company in Kathmandu can hire in Manila, Sydney and New York in the same quarter.
• Analytics — people data became something you could query, model and bring into the boardroom alongside revenue and churn.
• Generative AI — the most recent and most consequential shift: machines that can draft, summarize, screen, predict and coach at near-zero marginal cost.

Each shift expanded what HR could do. Together, they changed what HR is for. The move from local to global hiring, from intuition to analytics, and from manual processes to HR technology created the conditions for People Intelligence to exist.

Over the past decade, my work has taken me from traditional HR responsibilities into global talent acquisition, People Operations, organizational design, technology-enabled HR and, increasingly, AI. I didn't plan that arc — the work pulled me through it.

The ground truth behind it: more than ten years across six countries. More than 2,500 hires, from first employees to executive teams. Scaling organizations from 5 people to 350+. Running recruitment across borders and time zones, and People Operations across multiple countries at once — mostly inside technology companies, where the pace makes weak systems visible within a quarter. Implementing and rebuilding HR systems until they actually matched how the company worked. And, in the last few years, experimenting seriously with AI across every part of the employee lifecycle — including plenty of experiments that didn't survive contact with reality.

"Every stage taught the same lesson from a different angle: the companies that win are the ones that know their people best — and know it earlier than their competitors do."

I don't believe AI will replace HR professionals. I believe HR professionals who understand AI will redefine what HR can be.

This is not a defensive position. It is what the evidence from real teams shows. AI is genuinely excellent at the parts of HR that are high-volume, pattern-based and information-heavy. It is genuinely poor at the parts that require context, stakes, ethics and trust. The job of the modern People leader is not to pick a side in that argument — it is to design the boundary between the two, deliberately, process by process.

Get the boundary wrong in one direction and you automate away the judgment that makes HR worth having. Get it wrong in the other and you keep humans doing work machines do better, while your competitors compound the advantage.

AI isn't a concept I write about. It's a tool I work with.

I want to be clear about the difference between observing AI and practicing with it. Over the last few years I have used AI for experimentation and implementation across recruitment, HR operations, people analytics, performance management, learning & development and decision support. Some of those experiments became part of how teams operate. Many were discarded because the output was unreliable, the governance was unclear, or the human cost outweighed the time saved. That process — build, test, discard or adopt — is what moves someone from AI observer to AI practitioner.

These are applications I have used, tested or built with teams — not hypothetically.

• Recruitment — drafting and A/B testing job descriptions, generating structured interview kits, and pressure-testing scorecards before a single candidate is interviewed.
• Candidate screening — first-pass evaluation against a written rubric, with a human reviewing every borderline case and every rejection. Speed with explicit human oversight and bias checks.
• Talent intelligence — mapping where specific skills actually sit, what they cost in each market, and how competitors are hiring for them.
• Employee experience — answering the top recurring questions instantly, so HR's human hours go to the conversations that need a human.
• People analytics — turning headcount, attrition and engagement data into answers leaders can act on, in hours instead of weeks.
• Performance management — helping managers write clearer goals and better feedback drafts, which they then own and edit.
• Learning & development — personalized learning paths and on-demand coaching material, generated per role instead of per catalogue.
• HR operations — document drafting, policy Q&A, workflow automation and audit preparation.
• Workforce planning — scenario modeling for hiring plans: what happens to the org chart, the budget and the management load at each growth rate.
• HR decision-making — as a sparring partner: red-teaming a restructuring plan, stress-testing a compensation change, finding the argument I hadn't considered.

This part matters more than the list above, and it is where many AI-in-HR pitches go quiet.

• Final decisions on hiring, firing, promotion and pay. AI can inform; a person must decide — and be accountable for the decision.
• Employee relations and sensitive conversations. A grievance, a mental-health disclosure, a termination: these require presence, not prediction.
• Anything unverified about an individual. Models fabricate confidently. No output about a real employee is used until a human has checked it against reality.
• The ethical boundary itself. Deciding what should be automated is a values decision. It cannot be delegated to the thing being deployed.

The next generation of HR professionals won't be defined only by how well they understand people. They will need to understand data, technology, business, AI and organizational systems — while remaining deeply human. That is what the Future of Work actually demands: not a smaller HR function, but a more capable one.

That combination is rarer than either half alone, and it is exactly where the leverage is. A People leader who can read a churn model in the morning, red-team an org design at noon, and sit with a struggling manager in the afternoon is not a generalist. They are the connective tissue between the company's strategy and the humans who have to execute it.

The organizations that figure this out will treat their People function the way they treat their finance function: as an intelligence layer the business cannot run without — not a cost center that processes paperwork.

My goal is simple: to help build People functions that are as intelligent, measurable and adaptable as the businesses they support — without losing the human element that makes organizations worth building.

That is what people intelligence means to me. Not a dashboard, not a tool stack, not a job title. A way of running HR where every important people decision is informed by evidence, accelerated by technology, and made — still — by a human being who understands what is at stake.

HR didn't become obsolete. It became the most interesting function in the company. We just have to build it that way.