The Great Stink: What London Teaches Us About HR Infrastructure
In the summer of 1858, London nearly collapsed into its own waste. The city had grown to 2.5 million people, but the infrastructure beneath it hadn't kept pace. Sewage drained into the Thames. The Thames supplied the drinking water. Cholera killed tens of thousands. Then a summer heat wave concentrated everything into what history calls The Great Stink: a smell so overwhelming that Parliament suspended session and fled.
The crisis forced a question that changed the city forever: before London could become what it was trying to become, what did it need to build underneath?
The answer fell to a civil engineer named Joseph Bazalgette.
Bazalgette didn't patch the symptoms. He identified the structural gap and designed a solution at scale. Eleven hundred miles of sewers beneath the city. Waste redirected away from the drinking supply. Embankments along the Thames that buried the infrastructure and beautified the city above it. A modern city became possible because someone finally built the infrastructure layer it had been missing.
HR technology in 2026 is living in its version of 1858 London, and the AI moment is HR's Great Stink. The average enterprise runs a dozen or more HR systems, accumulated through mergers, legacy decisions, and years of point-solution sprawl. And now AI is being laid on top of it: agents that promise to answer workforce questions, automation that executes policy, and analytics that surface insight on demand.
The ambition is real. But underneath it, in most enterprises, manual work festers in pools, inconsistent definitions infect reports, and AI is making the contamination faster and more confident
People Ops’ Bazalgette Moment
The instinct is to fix the thing that's visibly broken. A weekly status meeting for the integration that keeps breaking. A better visualization for the dashboard nobody trusts. A different AI vendor when the last one hallucinated benefit plan details.
Each fix is rational in isolation. None of them addresses the underlying condition.
London did the same thing for decades before Bazalgette. The city built cesspools. It passed legislation requiring waste to be cleared from the streets. It added water filtration at individual pumps. Earnest, well-intentioned interventions, each aimed at a visible symptom. But the cholera kept coming and waste choked the streets because nobody had named the structural problem yet. The infrastructure layer didn't exist, and no amount of point fixes was going to create it.
The structural problem in HR isn't integrations. It isn't data quality. It isn't AI readiness. Those are symptoms of the same underlying condition: there is no control layer between HR's systems of record and everything sitting on top of them.
The systems of record hold the data. Analytics solutions restructure it for their dashboards. AI vendors pitch agents on top. But the layer in the middle doesn’t exist. There’s no control layer to govern how data moves between systems, how workforce metrics are defined, how policy is applied, and how AI is permitted to reason.
Forward-looking People Ops leaders are asking Bazalgette's question. Not "which tool do we add next" but "what do we need to build underneath before any of this works the way we need it to?"
That question is the inflection point. It's the moment the problem gets named correctly. Not as a collection of isolated friction points, but as a missing infrastructure layer. And once you've named it correctly, the path forward looks completely different.
In the next article, we'll look at what that missing layer actually is, what it governs, and why building it changes everything sitting above it.
Follow the story by subscribing here.