Microsoft Wasn’t Kidding

In the first edition of Margin Lives in Delivery, I came home from Directions NA with a clear message from Microsoft: Fix your data. Transform for AI. Empower the humans in the loop. Elevate ROI to the front of pricing discussions.

Well, they weren't kidding around. Success by Design has now evolved for AI-first delivery, and this isn't an "add an AI section to the Solution Design" kind of update. Agents become first-class components of the architecture. AI use cases and data readiness move upstream. Human-in-the-loop, agent boundaries and responsible AI become design decisions. Testing expands to evals, accuracy, safety and nondeterministic behavior. And the Solution Blueprint becomes a living artifact, continuously refined as the solution and the business evolve.

But underneath it all is a thread that hasn't changed. It's just become more important: the business outcome has to lead.

Not the software. Not the agent. Not the cool builds. "The why."

What outcome are we trying to achieve for the customer? What does the person doing the work today actually need? What friction are we removing? What decision are we improving? How much time, cost or risk does the automation eliminate? And is the return worth what we're asking the customer to spend to build, govern and maintain it?

Those questions start in the sales conversation and carry all the way through deployment.

Successful AI-first implementations start with presales and delivery architects working from the same outcome, from the beginning.

In an AI-first practice, presales qualifies the business case, not just the technology. Discovery protects the intended outcome while uncovering the data and context required to achieve it. Architecture has to drill into the business requirements to achieve the simplest structure of meeting the business expectations. PMs keep scope, decisions and expected value connected as the project moves. And at deployment, success goes beyond whether the agent works.

Did it produce the outcome we sold?

Practices that bring operational maturity and architectural discipline together from the beginning have a powerful advantage.

ModernOps is kicking it up a notch

I'm excited to share that Thomas White is joining ModernOps as Solution Architect.

Thomas has been doing this since Navision 1.2, which gives him an almost unreasonable amount of product history and solution architecture experience. (And makes him old. It's fine. He knows.)

But what matters to me isn't how long Thomas has been around. It's how he architects.

Thomas is a Solution Architect in the fullest definition of the role. He starts with the business problem, understands the process and protects the outcome before deciding what the technology needs to do. And he has always been disciplined about using MVP principles to solve the right problem with the minimal solution necessary to produce the intended value, rather than allowing "wouldn't it be cool if…" to turn into six months of customization.

That MVP discipline becomes even more important with AI.

Because AI gives us an extraordinary number of things we could build. A good solution architect keeps asking whether we should.

What's the value? What's the cost? What's the risk? Where does human judgment add more value than automation? And what is the smallest responsible solution that proves the business case before we expand it?

That's the architectural discipline I want sitting next to the ModernOps operational discipline.

I see the operating system around delivery — roles, process, governance, data, interfaces, margin and all the places organizational friction gets in the way of talented people doing good work. Thomas brings the architecture lens to make sure the solution stays anchored to its intended business outcome.

Same destination. Better vehicle.

Issue #1 was titled:

Microsoft gave you the directions. Your delivery model determines the ETA.

Three months later, it still holds.

Success by Design is evolving. Architecture is evolving. Roles, artifacts and testing are evolving. But the destination hasn't changed: create measurable business value for your customer.

And AI gives us a much more powerful vehicle to get there.

Now it's our job to make sure we actually reach the destination and aren't just doing burnies in the parking lot.

Welcome aboard, Thomas. Time to kick it up a notch.

Let's elevate.

— DGM