Building the Intelligent City Operating Model: How PX42 Turns City Data into Action

Charles Skamser

Welcome to Inside PX42. I’m Charles Skamser, PX42 Co-founder and CEO, and today we’re talking to city leaders about a very simple idea that sounds obvious, but it changes everything: becoming an AI-driven Intelligent City is not a dashboard project. It is an operating model. I’m joined by Catherine Spencer and Edward Hamilton. Catherine, Edward, good to have you both.

Catherine Spencer

Always a pleasure, Charles. And I’m glad you started there, because far too many cities think they are implementing AI when, in fact, well, they are just leveraging a chatbot and creating a shiny reporting layer and a colorful static dashboard. Helpful, perhaps, but not transformational.

Edward Hamilton

Quite. A dashboard tells you what happened. An operating model helps you decide what to do next, who owns it, and how quickly you can act before the matter becomes public and political.

Charles Skamser

Exactly. Most cities already have data. They’ve invested in departmental systems. Public Works has its tools, utilities have theirs, finance has another view, permitting another. The problem is not absence of software. The problem is leadership lacks one city-wide picture that supports coordinated action at speed.

Catherine Spencer

And the cost of that gap is bigger than people think. It’s not just inconvenience. Reactive operations show up in overtime, contractor surge, congestion, utility loss, permitting delays, emergency procurement, and frankly resident frustration. Staff work incredibly hard, but the city still ends up running on heroics.

Edward Hamilton

Yes, and heroics are expensive. Overtime is often treated as a labour issue, but it is usually a coordination issue. If work orders, staffing, routing, weather, events, and service demand are disconnected, the city compensates with buffers and extra hours. That solves today and worsens tomorrow.

Charles Skamser

I’ve said this to a lot of executives: if your teams are constantly firefighting, the city is paying twice. First in direct cost, then in the hidden cost of fatigue, delays, missed preventive work, and council headaches. Same story with traffic. Congestion is a tax residents pay in lost time, and cities pay in fuel, fleet wear, slower routes, and more high-visibility failures.

Catherine Spencer

Permitting is another one people underestimate. It looks like process, but it behaves like economic friction. When timelines are unpredictable, projects stall, businesses delay, housing slows, and public trust slips. Residents may not say, “Our operating model is fragmented,” but they absolutely feel the consequences.

Edward Hamilton

So the promise of an Intelligent City is not abstract. It is better service reliability, faster and calmer decisions, and stronger trust because leaders can explain what is changing and why, using evidence rather than anecdotes.

Charles Skamser

Right. Fewer surprises. Faster decisions. Clear accountability. And maybe most importantly, a city manager can move the conversation from explaining problems to debating priorities and investments. That’s a very different leadership posture.

Catherine Spencer

It also makes the programme defensible. If you frame this as “we’re buying AI tools,” you invite scepticism. If you frame it as “we are improving how the city delivers outcomes within a budget cycle,” that lands with finance, council, and department heads.

Charles Skamser

That’s the heartbeat of this episode. Don’t buy another disconnected tool. Build a measurable operating system for the city. In the next section, let’s make that practical and talk about what the City Health Layer actually does.

Charles Skamser

So let’s get concrete. When we say City Health Layer, we mean a 360-degree view of city operations with near-real-time visibility. Not just one department. Not just one dashboard. A leadership view of how the city is performing, where risk is building, and what to do next.

Edward Hamilton

And the crucial word there is “correlates.” Cities already have signals everywhere: service requests, work orders, asset performance, staffing signals, financial drivers, telemetry, external conditions like weather or events. Individually, each may be useful. Together, when correlated properly, they become operational truth.

Catherine Spencer

That’s where this gets valuable for executives. The City Health Layer is not simply showing more data. It is separating noise from what matters and translating it into the language leaders manage to: service reliability, response time, cost exposure, resident experience.

Charles Skamser

And it’s forward-looking. Once you establish a reliable baseline, the system can forecast near-term health, really the next 24 hours, based on patterns and leading indicators. That could mean seeing a likely 311 surge after a weather event and recommending that crews be pre-positioned before overtime becomes the default response.

Edward Hamilton

Or predicting congestion risk near a planned event and suggesting temporary operational changes and communications to protect mobility and emergency response times. Same principle in utilities: if telemetry and failure patterns indicate elevated risk, leaders can prioritize inspection and preventive action before there is an outage and emergency spend.

Catherine Spencer

That “before it becomes public” point matters. The best leadership use of a City Health Layer is not to admire the forecast. It is to recommend course corrections while there is still time to change the outcome.

Charles Skamser

Yes, and that’s why we talk about playbooks. If the system spots rising risk, it should help leaders test interventions. Add two crews? Shift inspections? Stage contractors earlier? Rebalance routes? Delay non-critical work? The point is to make decision support practical, not theoretical.

Catherine Spencer

And self-service matters here. Leaders should be able to ask in natural language, “What are my top three risks to service reliability tomorrow?” Then follow with, “What if we shift staffing in the morning?” They should not have to wait two weeks for an analyst, or a data science team that’s already oversubscribed.

Edward Hamilton

Quite right. In well-designed observability, the user does not need to know the plumbing. They need confidence that the insight reflects approved data, consistent KPIs, and current conditions. Then they can act.

Charles Skamser

And when that happens, leadership changes. You’re not managing by anecdote. You’re not debating whose report is right. You’re managing to one KPI spine, one shared picture of city health, and a clear line from signals to decisions to action.

Catherine Spencer

Which, by the way, is also how trust improves internally. Departments stop defending their own truth and start working from a common operating picture.

Charles Skamser

Exactly. Near-real-time visibility, cross-department correlation, early warning, and near-term forecasting. That’s the City Health Layer. Next, let’s talk about where cities should start if they want fast ROI and visible wins.

Charles Skamser

One of the biggest mistakes cities make is trying to boil the ocean. If you want fast ROI, start with high-volume, high-visibility domains where the data is reasonably strong and the outcomes are visible inside a budget cycle.

Catherine Spencer

Yes. Pick two domains, not ten. And choose them deliberately. You want a mix of operational value and political credibility. High-volume services like 311, field operations, traffic, facilities, and permitting are excellent first-wave candidates because residents feel the difference quickly.

Edward Hamilton

311 is particularly effective because it sits close to resident experience. When requests, work orders, inspections, and communications are connected, the city can forecast demand spikes, route work more intelligently, and reduce repeat contacts. Fewer unresolved tickets is not an academic metric. It is visible trust.

Charles Skamser

I love 311 as a quick win because it’s measurable and cross-functional. Same with field operations. Public Works, sanitation, parks, inspections, facilities—those groups already have budget tied up in overtime and contractor surge. Better forecasting and routing can reduce premium labour and missed service levels pretty fast.

Catherine Spencer

Traffic is another strong starting point because people feel congestion every day. If you can reduce choke points around events or construction and protect response routes, residents notice, businesses notice, and council notices. It’s one of those areas where a modest improvement gets outsized visibility.

Edward Hamilton

Facilities also tend to produce a clean ROI story. Lower utility spend, fewer reactive maintenance dispatches, and better control of peak demand are all measurable. Finance leaders rather like savings that show up on actual bills.

Charles Skamser

And permitting—look, if you can reduce backlog and improve cycle-time predictability, you’ve got a very defensible story on growth, affordability, and responsiveness. You don’t need perfection in year one. You need visible improvement with evidence.

Catherine Spencer

That’s the key. Choose two domains with strong data and visible outcomes. Traffic plus field operations. Facilities plus water-related operations. Permitting plus inspections. There isn’t one magic pairing, but there is a logic: start where the city can prove value quickly.

Edward Hamilton

And those quick wins matter beyond the operations themselves. They create budget-cycle ROI, which is the language leadership must defend. They also create credibility with council because the city can point to trend lines, interventions, and measurable results rather than promises.

Charles Skamser

Momentum matters. Early wins fund further modernization. They also build confidence across departments that this isn’t another pilot that dies quietly after a kickoff meeting and three steering committees. We’ve all seen that movie.

Catherine Spencer

We have, unfortunately.

Charles Skamser

So if you’re a city leader listening, your first move is not “What platform do I buy?” It’s “Which two domains can show measurable improvement fastest?” Once you have that, you can scale from proof to programme. And that brings us to how PX42 and UBIX fit into the stack you already have.

Charles Skamser

Let me be blunt about this: PX42 does not come in and tell a city to rip and replace systems that already work. That is expensive, disruptive, and usually unnecessary. Our approach is integration first.

Edward Hamilton

Which is sensible, because most cities already have a patchwork of systems of record and systems of action: GIS, permitting, ERP and finance, HR, work order platforms, 311, utilities telemetry, fleet, document systems, and various reporting tools bolted on over time.

Catherine Spencer

Exactly. The issue is not that those systems are worthless. The issue is they are fragmented. Leaders can see signals in isolation, but they cannot easily connect them into one decision-ready picture. That is where UBIX comes in.

Charles Skamser

UBIX is the AI-driven analytics and correlation layer. In plain English, it ingests and correlates data across departments and domains, combines business KPI analysis with infrastructure, application, and outside-influence analysis, and turns fragmented signals into business truth correlation and a business health dashboard.

Edward Hamilton

Yes, and that last bit is important. The observability concept here is broader than traditional IT observability. It is business health observability. So you are not merely watching systems; you are understanding how operations, workloads, external conditions, and service outcomes interact.

Charles Skamser

Right. Data ingestion and correlation, KPI analysis, outside influence analysis, predictive analytics, course corrections, remediation. That’s the chain. And because it sits above the current estate, staff keep working in the systems they already use.

Catherine Spencer

Which is critical for adoption. If a supervisor has to abandon the work order environment they know, or permitting staff have to jump into some brand-new analytics tool every day, adoption suffers. If insights flow into existing workflows, the city moves faster with less friction.

Charles Skamser

And one of my favourite parts is self-service. Leaders should be able to use natural language to ask, “What’s driving missed service levels this week?” or “What if we shift contractor capacity versus extending shifts?” They should not need scarce data science resources every time priorities change.

Edward Hamilton

Indeed. Self-service what-if analysis is not a luxury; it is capacity expansion for leadership. It keeps decision intelligence in the hands of the people accountable for outcomes.

Catherine Spencer

And it reduces dependence on bespoke analytics projects. That matters in city environments where time, budget, and specialist talent are all constrained.

Charles Skamser

So the model is simple: keep the systems you’ve invested in, add intelligence and coordination above them, and give leaders a self-service way to see risk, test options, and act. That’s how PX42 and UBIX fit the city’s existing stack. No drama, no rip-and-replace fantasy. Just operational lift. Now, of course, none of that works safely without guardrails.

Charles Skamser

I’ll say this as clearly as I can: if city leaders cannot defend how AI-supported decisions were made, the programme will stall. Guardrails are not optional. They are foundational.

Catherine Spencer

Absolutely. And when we say guardrails, we mean practical controls the city can explain in plain language: truth, policy, permission, and audit. Those four ideas cover most of what leaders need to make AI useful and safe.

Edward Hamilton

Truth means the system answers from approved sources and ties recommendations back to authoritative evidence. It does not improvise. If evidence is insufficient, it should say so, ask for clarification, or escalate to a human reviewer.

Charles Skamser

That’s a big one. AI must not “fill in the gaps” in a city context. If it doesn’t know, it has to say it doesn’t know. Or it escalates. That’s how you avoid bad recommendations, resident-facing mistakes, and the classic nightmare where somebody asks, “Why did the system say that?”

Catherine Spencer

Policy is the second rail. The system must follow city policy, ordinances, labour constraints, and workflow rules. Sensitive actions—resident communications, operational plan changes, dispatch triggers, policy interpretations—should require the right approval gates.

Edward Hamilton

Permission is equally important. AI should only access authorised data and respect decision rights. Not every user should see every signal, and not every recommendation should be executable by every role. Separation of duties matters, particularly in higher-risk workflows.

Charles Skamser

And then audit. Leadership needs traceability. What information was used? What rules were applied? Who approved the action? When did it happen? If you can’t answer those questions, you do not have city-scale governance.

Catherine Spencer

The lovely thing—well, lovely may be too strong a word—but the practical thing is that good guardrails actually accelerate adoption. Department leaders and frontline teams are much more willing to use AI recommendations when they know the system is constrained, explainable, and accountable.

Edward Hamilton

Quite so. Safety is not the enemy of speed. Undisciplined systems create hesitation. Governed systems create confidence.

Charles Skamser

And this is where PX42 is very deliberate. We implement guardrails as an operating layer from day one, not as an afterthought. Source controls, permission controls, truth enforcement, approval paths, audit logging. In some cases tiered risk modes as well, where higher-stakes workflows get stronger controls.

Catherine Spencer

Which is exactly what public-sector adoption requires: accountability, traceability, and clear decision rights. That is the foundation for safe city-scale use.

Charles Skamser

So if you’re evaluating AI for city operations, here’s the test: can it prove where the answer came from, whether it followed policy, who had permission, and what was approved? If not, you don’t have a leadership-ready solution. In our last chapter, let’s talk about the execution roadmap for becoming the model city.

Charles Skamser

Alright, let’s finish with execution. Vision is nice. Roadmaps matter. If a city wants to become a model Intelligent City, the sequence is pretty practical: pick the first two domains, stand up governance, launch a rapid start, and build adoption rhythms.

Edward Hamilton

And in that order, preferably. If you begin with vague ambition and no operating structure, you drift. If you begin with two concrete domains and a measurable charter, you can create traction rather quickly.

Catherine Spencer

The governance piece should be small and empowered, not bloated. City manager’s office, finance, IT, and leaders from the first-wave domains. Their job is to agree the KPI spine, clarify decision rights, and ensure truth and audit policies are in place.

Charles Skamser

Then launch a rapid start. Not a grand, eighteen-month architecture exercise. A working first wave that integrates priority systems, stands up the City Health Layer, and deploys the first playbooks. In traffic and field ops, that might be next-day workload forecasting and route rebalancing. In facilities, it might be energy stress forecasting and prioritized maintenance. In permitting, backlog visibility and bottleneck forecasting.

Catherine Spencer

And then you make it real through cadence. Weekly reviews with leadership to prioritise risks and decide interventions. Daily huddles in operational teams to act on near-term signals. Monthly performance checks to validate outcomes and decide what scales next.

Edward Hamilton

Those rhythms are deceptively important. They stop the programme becoming ceremonial. The city sees the signal, takes the action, and reviews the result. That closed loop is what turns analytics into management.

Charles Skamser

Exactly. And adoption has to be measured too. Are the reviews happening? Are playbooks being executed? How fast is time-to-action on priority risks? Are we actually moving overtime, backlog, cycle-time predictability, service consistency? If the answer is no, fix adoption before you declare the model broken.

Catherine Spencer

I’d add one more leadership point: report progress simply and consistently. A small set of stable metrics, clear trends, plain-language explanation of what changed and why. That’s how trust builds with council, staff, and residents.

Edward Hamilton

And, if I may sound slightly grand for a moment, cities that can measure health, forecast risk, and coordinate action with accountability will become the reference models others follow. That is the strategic prize.

Charles Skamser

That’s exactly right. Not because they bought more technology, but because they learned to run the city with better visibility, better foresight, and clearer accountability. If you’re a city leader listening, start with two domains, make governance real, move fast with guardrails, and build the cadence that turns this into how the city runs.

Catherine Spencer

Nicely said, Charles.

Edward Hamilton

Indeed. A practical agenda, not hype. Always refreshing.

Charles Skamser

That’s our goal. Catherine, Edward, thanks as always. And to our listeners, we’ll keep unpacking what practical AI leadership looks like in future episodes of Inside PX42. Until next time, take care.

Catherine Spencer

Thanks, Charles. Goodbye, everyone.

Edward Hamilton

Thank you both. Goodbye.