Tired of This Recurring Problem

It's a story I've seen play out far too many times here in Central PA, and I'm sure many of you, my fellow Digizens, have lived it, too. The conversation starts innocently enough: "We need to get serious about AI." Enthusiasm is high. The budget is approved. A team of talented engineers and data scientists—some of the best minds in our region, right here in Harrisburg, Lancaster, and York—are put on the project. They dive into the data, build incredible models, and come up with something truly innovative.

And then… nothing.

The project stalls. The polished presentation decks gather digital dust. What was meant to be a groundbreaking improvement becomes a cautionary tale whispered in the breakroom. The frustration is palpable. The engineers feel like their hard work was for naught. The business leaders wonder why they invested so much for so little. The effort feels like a massive exercise in rework, a cycle of starting over without ever actually finishing. We all get caught in this loop, pushing a rock up a hill only to watch it roll back down, over and over again.

We're so focused on the shiny new tool—in this case, AI—that we never stopped to figure out what we were actually trying to fix. We’re looking for AI to provide more efficiency, but because we haven’t measured the efficiency of the current process, it’s hard to measure the business impact that AI can provide. This leads to cases where the cost of improving efficiency is higher than the improved business value. We're so busy building the next big thing that we completely ignore the foundational processes that would allow it to succeed. We touched on this problem in a previous article, “The AI Secret Exposed In Central PA: It's Not About the Code, It's About the People,” where we highlighted that a significant percentage of AI projects fail to reach production for this very reason. This article is a summary of a video from the "Implementors" channel titled, "AI in the Real World: What Business & Education Leaders Are Actually Doing (and Learning) Right Now," where Don Demcsak from the Lancaster AI Symposium 2.0 was interviewed and talked about using Value Stream Mapping for AI projects.

There must be a better way.

Let's Define the Goal

We've all been there. A business leader comes to us and says, "We need to use AI to make our process more efficient." We nod, we smile, and we get to work. But the problem isn't the AI. The problem is what's upstream of the AI. Before we even consider which algorithm to use or what data to feed it, we need to understand the fundamental process we're trying to improve. This isn't just about applying a tool; it's about a small-scale Digital Transformation of how we approach problem-solving in the first place.

Why does this problem of stalled projects keep happening? It's because we're focusing on the solution before we've fully defined the problem. We’re presented with a symptom, "Our process is slow," and we jump to a solution, "Let's use AI to speed it up." But what if the process is slow because of inefficient hand-offs between teams? Or because of unnecessary approval loops? Or because the data itself is a mess? AI can't fix these things on its own. In fact, if you pour a powerful tool like AI into a broken process, all you get is a faster, more expensive broken process.

This is the equivalent of trying to improve a grocery shopping trip by inventing a new type of shopping cart when the real issue is the poorly organized list you use. The cart might be sleek and futuristic, but if your list is a jumbled mess of "milk," "bread," and "something for dinner," you're still going to be running back and forth across the store. The real problem isn't the shopping cart; it's the "what to restock" process itself. In our case, the real issue isn't the AI model; it's the lack of a defined, measured, and understood process for creating business value.

Goal Definition

Our goal is not to "implement AI." Our goal is to create a new, improved process that allows us to clearly understand, measure, and optimize a workflow before we introduce a powerful new tool like AI. We need a way to visualize the entire process—from start to finish—so we can see where the waste, the delays, and the rework are happening. This new process must provide a clear baseline so we can accurately measure the impact of any changes we make, including the introduction of AI. In short, we need to be able to answer the question, "Is this new tool actually helping us achieve our business goals?"

First Iteration

To make this a tangible goal, let's frame it as a user story, which I've found is one of the most effective ways to crystallize a problem and its solution.

As a business leader,
When I consider using a new technology like AI to improve a process,
Then I should be able to see a visual map of the current process and its inefficiencies,
So that I can make a data-driven decision about where to apply the technology for the greatest business impact.

Now that we know what we need, we can look for a tool. I've found that Value Stream Mapping is the simplest and most effective thing that will work here.

Value Stream Mapping isn't a new concept, but it's a powerful one that has found new life in the age of digital transformation. It's a lean management technique that allows you to visually document the flow of materials and information required to deliver a product or service to a customer. Essentially, it's a detailed, step-by-step diagram of your entire process, from the initial trigger to the final delivery.

Here's how it works and why it's the perfect tool for our AI problem:

1. It Forces You to See the Whole Picture: The first step in Value Stream Mapping is to stop thinking about a single part of the process and start thinking about the entire chain of events. For an AI project, this means mapping out every step, from the initial business request to the data collection, the model development, the integration into an application, and the final user feedback. You literally draw a map of the process.

2. It Identifies Waste and Inefficiency: As you map out the process, you're looking for what we call "waste." This could be anything from a redundant approval step to a delay in data delivery or a hand-off between teams that requires unnecessary rework. The beauty of the visual map is that these points of friction become impossible to ignore. You can see the time spent in each step versus the time spent in transit between steps. You'll often find that the real problem isn't the work itself, but the time it takes to get from one person to the next.

3. It Establishes a Baseline: Once you've mapped out the current state of your process, you have a baseline. You can now measure the total cycle time, the time spent on value-added tasks versus non-value-added tasks, and the number of hand-offs. This is the critical step we've been missing. Now, when you introduce an AI tool, you can go back to your map, update it, and compare the new numbers to the old ones. Did the AI actually reduce the cycle time? Did it eliminate a hand-off? With a Value Stream Map, you'll have a clear, data-driven answer.

4. It Fosters Collaboration: Value Stream Mapping is not a solo activity. It's a team exercise that brings together all the stakeholders in a process. The business owner, the data scientist, the engineer, and the end-user all sit in the same room and contribute to the map. This shared understanding is invaluable. It breaks down silos and ensures that everyone is working with the same information and towards the same goal. The data scientist understands why a delay in data collection is a problem for the business, and the business owner understands the complexities of model deployment.

By using Value Stream Mapping, we're not just finding a place to plug in AI; we're creating a robust, optimized foundation for any future technology implementation. We're turning a potential failure into a guaranteed success by making sure the AI we're building is solving a real, measured, and understood problem. This is how we ensure that our projects actually make it to production and deliver real value.

The Rollout and the Feedback Loop

Now, I know what some of you are thinking. "This sounds great in theory, Don, but my team is already swamped. How do I introduce another new process without a massive fight?"

This is a fair point, and it gets to the heart of why so many improvement attempts fail. It's not the technology or the process itself; it's the people. The human element of change is the single biggest hurdle we face. So, if we're going to use Value Stream Mapping, we need to do it right.

First, the key is a low barrier to entry. This isn't an academic exercise with complicated software. All you need is a whiteboard, some sticky notes, and a dedicated hour with the team. Don't frame it as "we're going to implement a new lean management process." Instead, say something like, "Hey, let's spend an hour drawing out how this process works today, from start to finish. I'm just curious to see all the steps." By focusing on discovery and understanding rather than judgment and change, you'll get far less resistance. You're not telling them their process is broken; you're asking them to help you understand how it works.

Second, you have to establish an active feedback loop. The Value Stream Map isn't a one-and-done document. It's a living artifact that needs to be revisited and updated. After you've identified a bottleneck and implemented a solution (like a new AI model), you need to come back to the map and measure the results. Did it work? Did it make things better or worse? This continuous improvement cycle is what makes the process stick. Without a feedback loop, the map is just a static drawing, and this is just another improvement attempt that fails to gain any traction.

The goal is to get your team to own the process, not just follow it. When they see their own suggestions and improvements reflected on the map and see the tangible positive results, they'll become advocates for the process, not just participants.

The Central PA Pulse

As we continue to navigate the ever-changing tech landscape, it's inspiring to see how our local ecosystem is growing and evolving. Here's a look at what's happening right here in our backyard.

Regional News

• Lancaster Cultivating a Bustling Tech and Business Ecosystem: The city of Lancaster is making waves with its thriving startup scene, attracting innovators and entrepreneurs from across the region. With initiatives aimed at fostering growth and collaboration, Lancaster is quickly becoming a hub for technology and business, showing that our region is a fertile ground for new ideas and ventures.

• Local Officials Are Preparing For A Wave Of A.I. Data Centers In York County: York County is bracing for a new wave of development as officials plan for the influx of AI data centers. This points to the growing demand for local infrastructure to support the very technologies we're discussing, making it a critical topic for our community's future.

Local Events

• J-Dammit 2025: Mark your calendars! The J-Dammit 2025 conference is coming to Harrisburg University. This event is a great opportunity to connect with fellow Digizens, learn about the latest trends in technology, and share your own experiences and insights. You can find more information and register at the event's website.

What's Your Problem?

That's one problem down. Now I'm curious, what's the recurring problem you're tired of? Is it a project that's stuck in a loop, a process that's just not working, or a team dynamic that needs a change? Send me a note. Maybe we can figure out a "Digital Transformation" for it in a future post.

Social Media

Digizenburg Dispatch Community Spaces

Hey Digizens, your insights are what fuel our community! We've been diving deep into the world where AI meets BI, and we know many of you have firsthand experiences and brilliant perspectives to share. Let's keep the conversation flowing beyond these pages, on the platforms that work best for you. We'd love for you to join us in social media groups on Facebook, LinkedIn, and Reddit – choose the space where you already connect or feel most comfortable. Share your thoughts, ask questions, spark discussions, and connect with fellow Digizens who are just as passionate about navigating and shaping our digital future. Your contributions enrich our collective understanding, so jump in and let your voice be heard on the platform of your choice!

Facebook - Digizenburg Dispatch

LinkedIn - Digizenburg Dispatch

Reddit - Central PA

Our exclusive Google Calendar is the ultimate roadmap for all the can’t-miss events in Central PA! Tailored specifically for the technology and digital professionals among our subscribers, this curated calendar is your gateway to staying connected, informed, and inspired. From dynamic tech meetups and industry conferences to cutting-edge webinars and innovation workshops, our calendar ensures you never miss out on opportunities to network, learn, and grow. Join the Dispatch community and unlock your all-access pass to the digital pulse of Central PA.