Build a Real-Time Dashboard Without New Systems

"How do you control all this? How do you know where you are, what's missing?"

That question came from Alfredo, a construction company owner with 30 people and 15 active projects. He was asking what every CEO asks: How do I know what's actually happening in my business right now, not at the end of the month when the finance team closes the books, not next Tuesday in a status meeting, not after someone sends me an email.

Most executives get their answer the same way: they ask someone. The project manager, the sales manager, the operations person. Or they pull up Excel. Or they do a walkthrough and try to piece together a picture from what they see and what they remember.

That's not visibility. That's assembly by committee.

Every mid-market company has this problem, and it's why we always recommend consolidating data before adding AI. Your data is in QuickBooks, Salesforce, Sage, maybe an old ERP, definitely a spreadsheet or two, possibly a tool your team built in Airtable when they got frustrated with the main system. Nobody thinks the numbers in one place match the numbers in another. The CEO spends Monday morning asking people for status instead of running the business. And when something breaks, a delayed project, a lost deal, a vendor invoice that's been sitting unsigned for three weeks, nobody catches it until it's a problem.

The real issue is not that you don't have tools. It's that your tools don't talk to each other, and there's nobody home to translate between them.

A command center, the second layer of the AI Operating System, is not just another PowerPoint with charts. It is a live view of the one thing that matters: Can you run your operation without asking people?

Revenue and margin tracking. Current month revenue, margin, and trend against forecast. For a construction company: contract values, cost status, and whether projects are on budget. For a legal firm: billable hours, realization rates, and which clients are profitable. Current, not historical.

Project or deal status. Traffic light view: which projects are on track (green), which need attention (yellow), which are broken (red). Red triggers an alert. The CEO sees it before the project manager does.

Team utilization. Who is allocated to what. How many billable hours are scheduled this month. Whether the bench is full or you have capacity. Particularly critical in service businesses where people are the product.

Cash flow. Outstanding invoices. Upcoming commitments. Days sales outstanding. A health check that happens daily, not at month-end.

Alerts. An invoice unpaid past 60 days. A project slipping 10% off schedule. A sales opportunity aging more than 90 days. The system flags it; someone acts.

The specificity depends on your business. An architectural firm cares about billable hours and client profitability. A CPG distributor cares about inventory movement and regional performance. A financial services firm cares about assets under management and advisor capacity. The framework is the same. What gets built depends on your operation.

You already have systems. QuickBooks is handling the accounting. Salesforce is tracking deals. Your ERP is running operations. The last thing you want, and need, is to rip and replace. You need someone to make the systems you have work together.

That's the function of a live dashboard. It connects to your systems through APIs, automatic data connections between systems. QuickBooks, Salesforce, Google Sheets, custom databases all have an API. Some are polished and documented. Some require a little work to find the data you actually need. But the systems themselves don't change. The dashboard pulls data automatically, updates it hourly or by the minute, and presents it in one place.

One distribution company came to me asking for exactly this: "Give me my Excel pivot in Power BI, intelligently." They were running their entire operation from a spreadsheet. We built a dashboard that pulled their data from their on-prem ERP, their inventory system, and their order book. In 6 weeks, the real-time view of inventory, orders in flight, and revenue forecast replaced the Excel file entirely. They still use Excel for scenario planning. But the operational reality now lives in the dashboard.

The key detail: I did not replace the ERP. It is still their system of record. The system did not change. The visibility did.

This is standard for mid-market companies with mature operations. You have too much data, too many tools, and too much institutional knowledge built into what you're running to justify a full replacement. You need translation, not replacement.

Alfredo's company. 30 people. $6 million annual revenue. 15 active projects. He ran the entire business from a spreadsheet, a 15-tab Excel file. Every Thursday, he or his project manager would update it: cost to date on each project, change orders approved, materials ordered, subs scheduled, estimated completion.

Here's what was broken. Alfredo did not know on Monday whether a $400K project was trending over budget. He found out Thursday when the spreadsheet was updated. By then, the cost overrun was three days older and harder to address. Second, the knowledge lived with one person. If that person was sick, or went on vacation, or left the company, nobody else could update the numbers. Third, the numbers had no context. A project was 30% over budget. Was that because of material costs, labor, or schedule delays?

We built a live dashboard in 6 weeks as an AI Foundation Build. The data source was the same: the project accounting system and the materials order log his accounting person already maintained. We did not change those. We added an automated pull that refreshed hourly. Now the dashboard showed current cost and budgeted cost for each project, cost variance by line item, schedule status, and alerts when a project exceeded budget by more than 10%.

Alfredo now logs in at 8 AM and sees the status. He does not ask for it. He does not wait for Thursday. When a project goes red, he has three days to act instead of waiting for a spreadsheet update. And the system is not on one person's laptop, it lives on his dashboard.

That's what a real-time dashboard does. It closes the gap between when something happens and when the person who can act finds out.

The cost and timeline depend on how many systems you're connecting and how clean your data is.

Simple dashboard (1 to 2 data sources, mostly clean data): fixed-fee build, 2 to 3 weeks. You're connecting one ERP and one CRM, or maybe pulling from accounting software and a Google Sheet. The data is structured. The transformation is straightforward.

Mid-range dashboard (3 to 4 sources, some data cleaning): fixed-fee build, 3 to 4 weeks. You're connecting an ERP, a CRM, accounting, and a project management tool. Some fields need mapping. Some data is stored differently between systems.

Full command center (5+ sources, complex logic): fixed-fee build, 4 to 8 weeks. You're building the central nervous system for your operation. Maybe you have an old ERP, a newer CRM, project management in a third tool, accounting in a fourth, and ad-hoc reporting in Excel that somebody's been maintaining for three years. The data architecture work is substantial. But the payoff is also substantial, you get true visibility instead of a patchwork.

These are fixed-fee engagements. You know the price before we start. If scope expands, you want to add a fifth data source, or the data turns out to be messier than expected, we talk about it. But the default is a fixed fee and a timeline.

A dashboard is not magic. It is not a substitute for good operations. If your business is chaotic, a dashboard will show you a real-time view of chaos. What it does do is collapse the lag between when something breaks and when you know about it. It moves visibility from async, email, meetings, spreadsheets updated once a week, to live.

For a 30-person construction company, that's the difference between catching a cost overrun three days after it starts and catching it six weeks later at project close-out. For a legal firm, it's the difference between knowing your attorneys are underutilized this month and realizing it in Q4 when you've already missed your revenue target. For a distribution company, it's the difference between knowing inventory is low and finding out when you miss a customer order.

Most of my clients started exactly where you are. They knew their data was scattered. They knew they were not getting answers fast enough. They didn't know how to fix it without hiring a data person or buying an enterprise platform for $500K. A dashboard, built right, costs a fraction of that, takes a few weeks, and you own the whole thing.

The questions to ask yourself: How much does it cost you in a month when you don't know something until it's too late? How many decisions would change if you knew the numbers live instead of at the end of a meeting? How much time does your CEO spend in status meetings instead of running the business?

Start there. That's the number you're comparing against the fixed-fee build for the dashboard.

If this sounds like your situation, data scattered across systems, visibility by committee, gaps between when something breaks and when you find out. take the free AI Ops Assessment. It will ask 10 questions about your current operation and your data architecture. At the end, you'll get a score and a specific recommendation about whether a dashboard is the right first step, or whether you need to fix your data layer first.