Real-World AI Use Cases You Can Deploy Today| | Ben Vollmer, RSM US Product leader

Samuel:
I’m really excited to have you on the show, Ben. Welcome.

Ben Vollmer:
Thanks for having me, Sam. It’s gonna be fun.

Samuel:
So I’m curious, what’s Power Factory actually?

Ben Vollmer:
So Power Factory is, if you think about it, our low-code, no-code offering inside of RSM. RSM is the fifth-largest tax and audit consultancy firm in the world. In the U.S., we’ve got 20,000-plus employees. We’re prolific. We actually are one of the biggest Dynamics 365 consultancies in the world. Power Factory is our offering just to do Power Apps, Power Automate, Copilot Studio — all of the Dynamics 365–adjacent products, the rest of the BizApps platform that is not D365.

Samuel:
And AI is now embedded inside everything — Power Platform, everything Dynamics, right? It’s just basically everywhere. And you’re in the field, right? You’re working with the customers, you’re working with some of the biggest customers in the U.S. So from the field, from what I heard, there’s a lot of noise right now created by the whole tech market, and it’s hard for leaders to know what’s truly available versus what’s just hype.

I’m pretty sure you hear about it as well. So from your perspective in the field, what should organizations start implementing today, and what belongs more in the one- to five-year horizon?

Ben Vollmer:
For me at least, there are two axes here. Let’s think about this from two axes. There is the technological axis — is this technology ready for me to actually adopt and do something with? A lot of stuff is changing very, very quickly. Is this ready to adopt? The second part of this, though, is what’s my organizational readiness to do this? I’ve seen some projects where we’ve had phenomenal ROI, but the organization wasn’t ready to adopt that area of technology.

And so what I’d say is, I would look at implementing what’s both technologically available and technologically sound for your use case, as well as what your organization is actually ready to adopt and ready to consume. Now, I’m talking about an organizational level, not a personal productivity level here, Sam.

Samuel:
It made me think of the BXT framework — the framework I’m using at Microsoft. I’m pretty sure you’re familiar with it — you know, Business, Experience, and Technology. So which one of these two you mentioned — technology or readiness — are you seeing most companies on?

Ben Vollmer:
Most companies I’ve seen — I think there are a few things. I’ve seen Skunk Works projects where we take something in a corner, try to build something in an innovation lab, and push it out to the rest of the company. That gets you a cool factor, that gets you some learnings, but that’s not truly an enterprise deployment. That’s kind of a “let’s go see if this thing works or not” and “let’s go kick the tires.” I think that’s one deployment we see.

Then I see kind of personal productivity — how to make myself better. You see that in M365 Copilot — “Summarize all of the Teams chats I missed yesterday,” “Give me a list of all the appointments that are upcoming today and who they’re with.” That’s about making me better.

And then you have team productivity — team automation. We’ve been doing a lot of stuff in the team realm, actually. The team realm, for me, is where our customers see the most value — things like invoice processing. We have an agent that handles accounts receivable for a large customer, where the supplier actually emails in and asks a question like, “Where’s my payment in this invoice app?” Instead of going to a human to look it up, find the invoice, and everything else, the agent reads the emails, looks in the accounting system, and replies back with the status of that invoice automatically.

That is where we see customers seeing a ton of value — in the team level.

But underpinning all this is the data. We talk about the Microsoft Graph, we talk about making sure your data is clean, making sure your data is ready. So I see customers starting with “get my data ready,” “get my team ready,” or “get my people ready.” You’ve got to pick which one of those three — or which combination — you want to blend into.

Samuel:
Interesting, because I think data is pretty much the foundation of everything AI right now — which you just described with the invoice processing. It’s basically agentic AI.

Ben Vollmer:
Well, agentic AI by itself — it is agentic AI, but it’s agentic AI blended with other things in the AI family. So OCR — like the invoice comes in, we recognize the invoice. I think one of the unsung heroes of the Microsoft ecosystem — you’ve seen that cartoon with the building supported by one open-source project — or the world’s financial system being supported by Excel. To me, AI Builder is that thing for Microsoft — the unsung hero of the product suite.

Samuel:
On that link about adoption, you shared with me the Gartner adoption timeframe. I found it really interesting — seeing what’s on the horizon, what’s really available today, what will be in five years, and where organizations should focus. So can you walk us through their findings and how you see that in practice with customers? Are you using this with customers?

Ben Vollmer:
I use it with customers every day. AI has a lot of “magic” being sold that’s not grounded in reality. There’s a lot of “we’re doing this” — and everybody’s doing it. It’s not just one person; everybody’s doing it.

I go back to those humanoid agent robots that were shown one time — and it turned out there were people in a back room actually controlling them. You see a lot of that “don’t look behind the curtain” type of stuff. So when I talk to customers, I ask — where’s the adoption value at? Where’s the adoption?

Computer vision, for example, has been around for a long time. Think about the power of being able to take a picture and describe it. We’ve done it for a volunteer organization where we wanted to understand how full the truck was after every donation pickup. Computer vision takes a picture of the back of the truck and says, “You have approximately 14 feet left in this truck.”

Samuel:
Is it accurate?

Ben Vollmer:
Yes. It’s surprisingly way more accurate than I expected. I wouldn’t use it where I need exact measures — feet, inches, millimeters, centimeters — but to get volume, it’s pretty good.

Edge AI — think about all the video in the world. You’ll be able to use video, for example, on the edge — that’s huge. I think we’re seeing big benefits there. Honestly, computer vision and document intelligence have probably been the number one areas where my customers see value.

Samuel:
If I go back to the Gartner adoption timeframe — at the beginning of adoption, we have computer vision and document intelligence. What’s on the two-, three-, four-, five-year horizon?

Ben Vollmer:
You’re going to see composite AIs. We’re seeing it now with MCP servers — A2A — where you have robots talking to robots. One of my favorite jokes right now is that I swear, most of the RFPs I see are generated by AI LLMs, and most of the answers to those RFPs are also written by LLMs. So how soon before those two just talk to each other and negotiate things — and tell us how much we should charge for deploying something?

Samuel:
Just for our audience — A2A stands for agent-to-agent, basically agents talking together. And MCP is Model Context Protocol, which is kind of the USB-C of AI — letting your software connect with agentic AI. Am I right?

Ben Vollmer:
Exactly.

Basically, think about Power Automate on steroids. Instead of having to build a system that encompasses everything, you can hand it off, let it do its thing, and it returns the results. It’s more of a macro service than a microservice for agents.

Samuel:
So on a close horizon, we have document intelligence and computer vision. In the next two to three years, we’re looking at agent-to-agent MCP servers that will take more places.

Ben Vollmer:
I think we’ll see them in more places. We’re seeing them come up pretty quickly. We’re seeing them come out, but again, the protocols are changing, and the use cases are changing. I’ve seen some use cases for them, but it hasn’t been broad spectrum yet — it’s been very segmented.

I think we’re starting to see more intelligent applications. James Phillips, who headed BizApps for a long time, used to say we have “forms over data.” That’s been the context since computers began. But what about when I start filling the form out and it says, “Hey, here’s the data,” and starts pre-filling or changing the form design based on what I’m entering?

You can do that today via coded applications, but not yet via free-form intelligent applications. I’m excited to see that.

We’re also starting to see some decision intelligence come out — like, “Should I do X or Y? Should I do A or B?” Like, “What should I do here? Help me, guide me.” Researcher, for example, is a great tool for that. It’s helping provide decision intelligence. But I think we’re still a couple of years away from people actually looking at data, trusting that data, and understanding it.

Samuel:
When you refer to Researcher, you mean Microsoft 365 Researcher, which is the deep reasoning model, am I right?

Ben Vollmer:
Correct. There’s also one in Cloud, one in Gemini, one in Copilot. There are a whole bunch of them out there. The Microsoft one, I think, is really, really good. I’ve actually had pretty good luck with the Gemini one too, believe it or not — especially around more product management functions. I think Gemini was written by product managers for their product management team, and it does some product management stuff really, really well that I enjoy using it for.

Samuel:
Yes.

It’s okay.

Ben Vollmer:
But those are things, though — even generative applications. Like, generative pages just came out in Power Apps. You’ve got things like Lovable. People like vibe coding — I’m not sure if “vibe coding” was even a word two years ago. Those are things that are going to change the outlook and the view of the world, I think, as we see it.

Samuel:
I totally agree. So we have our one-year horizon, two- to three-year horizon. What’s on the five-year horizon? Like, what should people start looking at, but not necessarily investing heavily in right now? Because I think MCP and agent-to-agent — you should start exploring that. But then there are those other techs that are promising a lot and will probably be here in five years. You should just maybe start looking at them, but not investing yet.

Ben Vollmer:
If I think about five years out, we’re going to be in a world where there are two types of AI. There’s going to be very broad, general AI — I’d put M365 Copilot in that category. It’s broad and wide. And then you’re going to have very micro-niche, micro-vertical, very targeted AIs for doing specific things.

Like, if I want to track the number of weather storms in North America, there’s going to be an AI that does that. If I want to do something else, there’s going to be another very micro-niche AI.

So, when we think about models going forward, I think the foundational models will be broad and horizontal, and then you’ll have very small micro-vertical ones. As an organization, you’ve got to prepare for how to make your broad, horizontal models snap into those micro-verticals as needed across your business.

So those foundational models might be: here’s my foundational model for my warehousing team, my law team, my supply chain team, my FP&A team.

Samuel:
That’s an interesting concept, because right now most foundational models are generalistic, right? Do you think those models will run on the edge — meaning they’ll run on your device instead of needing the cloud?

Ben Vollmer:
Mm-hmm. I think I’m interested to see what happens with SLMs — small language models. OK, so instead of large language models, give me some small ones. I want to see what happens when we can run an entire AI — an LLM — on my phone. I think those will be a big part of it.

Samuel:
What is that, Salman? Yes, OK.

Ben Vollmer:
LLMs are only a part of what people look at. There are things like machine learning. For me, most of what we call AI is actually ML.

But how do we get an ML model that does exactly what I have to do? That could run on the edge, on your device, or in the cloud. I don’t think that really matters — it should just be transparent to the end user where it sits.

Samuel:
Yes, right now everybody referring to AI is almost always referring to a large language model — so GPTs, to make that simple. But the truth is, there are other models out there that are very powerful. You talked about document intelligence and computer vision — those are machine learning models, right? They’re not LLMs; they don’t output text. I think a complete system will use a bit of everything — an LLM and other machine learning models.

Ben Vollmer:
Well, if I look at the projects we’ve done to date that have been the most successful, they’ve been a combination of models. We never use just an LLM. We use ML and LLMs in conjunction with each other — or vice versa. You have a set of checks and balances. This thing does it, and the other model says, “Yeah, that looks right.”

Those foundational models are really going to be critical to figure out. I’m most excited to see the micro-vertical side of things — like, where does ChatGPT for financial services institutions that are B2C-focused in Canada come from?

Samuel:
Interesting. Will there be enough training data to train those models? How do you see this specialization of those models happening?

Ben Vollmer:
I see that happening because — think about it — there are software companies. I mean, again, you worked at Microsoft for a while. Microsoft’s massive, right? Microsoft is a big fish in multiple big ponds.

Samuel:
Yes.

Ben Vollmer:
There are a ton of medium fish in medium ponds and small fish in small pond applications out there. One of my friends runs an EAM firm that does almost exclusively nuclear power generation plants. It doesn’t get much smaller than that.

But even in the Microsoft ecosystem, there are partners like RSM, for example. We have a food vertical we go after — dairy producers, fruit producers, production co-ops. If I said “production co-ops” to the average Microsoft employee, they’d probably have no idea what I meant. From the vertical view to the micro-vertical view gets lost sometimes. But there’s a ton of money and software being developed in those small micro-verticals. I think that’s where those models are going to come from.

Samuel:
Speaking of all those kinds of areas where AI can be applied — where do you see leaders that are actually realizing measurable value right now? Like, not in two, three, or four years — right now — with AI? And I’ll say AI agents, but let’s include a mix of different types of AI.

Ben Vollmer:
So agents, for me — I don’t think we’ve gotten to the view yet of fully autonomous agents. We’re not there. Even when we build systems, I want to put a human somewhere in the loop to validate data. We’re just not there yet. If somebody shows you an autonomous robot, I’d love it, but I haven’t seen one yet.

So what we generally do for our agents is put a human somewhere in the process to check it. When I look at where we can reduce the amount of human effort to do a project, that’s where the value is.

When we deliver something, it’s generally a combination of OCR and other things. The amount of paper in this world still shocks me — the amount of paper produced still absolutely shocks me. So we’re seeing value from taking paper, converting it into documentation or data, and passing it on.

I’m seeing a lot of value from our customers right now. Most of the Copilot Studio stuff we’re seeing on the public side is from B2C, but we’re seeing a lot of internal employee value — things like, “I’m running Jira, I’m running ServiceNow, I’m running an ITSM solution — how do I open up a ticket?” How do I make that process easier for my teams?

But I think the number one thing we see value in right now is dealing with documents — either incoming or outgoing — and summarizing data. We just did a project last week where we summarized shift data. When shift one finishes their job, they have to tell shift two what happened during their shift.

Samuel:
That’s great.

Ben Vollmer:
OK, so we summarize what happened during shift one so shift two doesn’t have to read 900 reports. Those are some great use cases. But the biggest thing we see value in is helping our customers determine — is this personal productivity (does this make Sam better?), does this make the solution engineering team at Microsoft better, or does this make all of Microsoft better?

We break it into those three buckets. There’s value in all three, but you’ve got to choose which one you’re going after.

People saw ChatGPT when it came out and said, “I love it! I want to roll this out to my whole organization.” Yeah, but you’re doing personal productivity. For organizations, you should really think about automating roles and tasks inside the company.

So there’s a lot of value we’re seeing — mostly around handling things that are human-intensive. Think about how much work it takes to open a document, read it, flip through it, match invoices — all that kind of stuff. It’s mind-numbing work. I don’t know how hard you are, but I don’t think anyone doing that says, “This is the best thing I’ve ever done in my life.” But it’s there.

Samuel:
So what are the top three departments where you see fast ROI adopting AI solutions? I’d suppose finance is one, based on what you just said — you know, using OCR for invoices, etc.

Ben Vollmer:
I’m always a little bit hesitant to say which department, because it depends. Every customer’s different. What we could call a finance function might actually sit in a field operations role. But I think the biggest ROI comes from the ability to look at large pieces of data and guide someone to a conclusion.

For example, a project we’re doing right now reads contracts and creates work order tasks from those contracts. That’s a boring job — I don’t know about you, but reading contracts isn’t exactly my favorite thing. But mapping that to work orders is huge.

So I think the biggest areas will be admin-heavy work. I was doing some research earlier today — salespeople spend over 70% of their time doing administrative tasks.

Samuel:
Yes.

Ben Vollmer:
And an interesting fact I learned from that: 99% of salespeople, when they’re given more time free from administrative tasks, use that time for more selling — which means 1% play golf.

But that’s a huge stat. Sales is often the highest-paid role in many organizations — so can we lift that burden there?

I also think any area where it’s the “cost of doing business” — like regulatory, financial, or risk compliance — is a great fit. Humans are really bad at the sunk cost fallacy.

You ever had an old beater car? You keep putting money into it — “I just put seven grand in it; I might as well put in another two to keep it running.” The truth is, at that point, the car’s shot. Put a bullet in it — move on.

Samuel:
What, is the sunk cost fallacy.

Transcript cleaner said:

Ben Vollmer:
So the sunk cost fallacy is: “I’ve already put this much money in, I’ll just do a little more.” When you do that, you avoid risk — and risk is a value measure. So if I can figure out a way to have AI look and say, “Hey, this is risky…”

Think about a statement of work for a consulting firm — “Hey, this is not scoped properly.” Or a grant — “I didn’t meet all the conditions in the grant, so I could lose that grant money.” Those are all areas where I see AI really helping — providing some guardrails where humans aren’t as strong.

Samuel:
So there’s not necessarily a specific department. It’s more, if I rephrase you, where there’s a large piece of data that brings you to make a decision — like the examples you just gave: using it in sales, revalidating scopes of work, contracts, etc. You need to look where there’s a lot of time spent reviewing data or information that can technically be passed to an AI — an agent or not — any kind of AI that removes the burden of having to do this manually, so you can spend more time on something meaningful.

Ben Vollmer:
One of the projects we did is a case from the Microsoft side. They had three or four employees, give or take, who were reviewing records one at a time. They’d been doing it for about 18 months and were only halfway through the process. The AI did about 95% of the records, and they finished the rest in six weeks.

Now, they didn’t fire those four people. Those four people now have jobs where they actually go do the value work — figuring out the rest of the process. Instead of manually reading records and updating things slowly, we did it much faster.

Samuel:
Now, you talked about it at the beginning — adoption is, I think, one of the biggest barriers to using AI. Getting non-technical people to first trust the AI, bring it into their daily workflow, and find new ways of working with it. How do you see organizations overcoming that challenge? Have you seen a lot of pushback from end users?

Ben Vollmer:
We have — I’ll tell you a story. When I was at Microsoft, we developed a tool that people called AI — it’s long math, that’s a whole other conversation — it’s Dynamics 365’s Resource Schedule Optimization, or RSO. One of the problems we had with RSO was that dispatchers didn’t trust it to make the same decisions they would make.

What we did was provide a simulation function where the dispatcher could see how RSO would model the day. They could accept it or reject it.

Think about yourself right now. I’m probably a little older than you, Sam, so I remember when the first GPS units came out — the TomTom or the Garmin you’d stick on your dash. When those first came out, I remember validating what it was doing versus where I wanted to go. Eventually, I started to trust it. And now you pick up your phone, type a company name, hit Go, and blindly follow it.

Samuel:
I sometimes joke that I don’t even know where I am anymore because I just follow the GPS — and honestly, it’s not really a joke.

Ben Vollmer:
Exactly. My parents live next door to me. I have two boys — 22 and 18 — and they sometimes take my parents to doctor visits. My mother drives them nuts. She’ll say, “OK, go down to 14th, take a left, then right on Third Street.” And my boys are like, “No, Grandma, what’s the address?” She says, “I don’t know.”

They lock horns because my boys want an address, and my mom only knows how to get there.

Now, think about how we bridge that organizationally. When Garmin and TomTom first came out, you monitored the GPS — you watched where it was going and decided whether to follow it. Now, you just hop in the car and blindly follow Waze or Maps. Every time I ignore Waze, I end up stuck in traffic.

So I think it’s about showing people how this empowers them. There’s fear around AI replacing jobs, but it’s not about replacing jobs — it’s about replacing jobs where people don’t use AI. The battle isn’t AI vs. humans. It’s humans who use AI vs. humans who don’t. That’s where the real difference will be.

Samuel:
So this will come with time — and people not choosing it will probably fall behind. That’s what you’re saying, right?

Ben Vollmer:
Exactly. You’ve got to ease people into it. Think about GPS again — we didn’t go from paper maps to fully trusting GPS overnight. That transition took a couple of years.

We have to do the same with humans — we can’t just blindly replace. Can the computer do the job? Maybe yes. But does the user trust it? Usually not.

With something like RSO — say you have 10 technicians with 10 appointments a day — I forget the exact number, but it’s something like 14 trillion to the 10th possible combinations. My wife and I can’t even decide where to go to dinner — that’s one decision, let alone millions. You have to show humans why this is the better choice.

Samuel:
Do you have a specific framework you’re using with customers?

Ben Vollmer:
We have a readiness framework we use. RSM is interesting — much like Microsoft. Microsoft isn’t one 800-pound gorilla; it’s 801 one-pound gorillas. RSM is like 800 quarter-pound gorillas.

We have an Office of Change Management with a framework: if you’re a new change manager, here’s how you do it. We have an Office of Risk Management: here’s how you handle risk. We have a deployment team with a security-oriented methodology.

It all depends on where the customer is in their journey. My team’s methodology focuses on three things: organizational readiness, ROI and value, and whether the technology can support the decision the customer made. We actually assign numeric values to those, and those drive how projects get done.

Samuel:
To help people adopt AI, I suppose empowering them will be helpful — letting them experiment. Like I’ll preach my own church here: giving them M365 Copilot, for instance, to start experimenting — letting them create agents themselves. That helps them see the value. It’s like going from paper maps to GPS.

Have you seen real-world examples where this democratization of AI — using low-code platforms or out-of-the-box products like M365, Agent Builder (now renamed Copilot Studio Lite) — has had a tangible impact on productivity and innovation? I’m pretty sure you have. Can you share some?

Ben Vollmer:
We have. I think a few things have happened here. One is getting people to trust the data sources. People fear the unknown — if there’s a black box of data that just goes in and out, they get scared. So the first thing we do is help people understand how it works — how you got from here to there. That’s critical.

As for the agents — consulting has been flipped on its head. Most of what my team does now is enablement. We’re basically enablement and “backlog as a service.” I hate that term, but that’s what it is. We help customers understand how to do this themselves.

Think about it like fishing — sometimes you get a fish that’s too big to pull in alone. So you call others over to help drag it into the boat. It’s the same idea: customers come to us with a backlog — “We need these 52 things built for our live product” — and we help them deploy it.

Consulting used to be: “Sam, what do you want to see? We’ll do the discovery and deliver the project waterfall-style.” Now it’s: “Here’s how you make yourself better. Here’s the right tool for the right job — don’t use Power Pages for this, don’t use Canvas apps for that, use this one instead.”

It’s all about enabling customers to build their backlog and manage it properly.

Samuel:
I think there’s this default belief — because of the way it’s marketed — that it’s so easy to create a website using Power Pages or an agent using Copilot Studio. The truth is, I think it still requires very specific skills and knowledge, especially with the speed things are evolving.

Do you see customers being able to manage all this by themselves, without help from RSM?

Transcript cleaner said:

Ben Vollmer:
Thank you.

I have customers who do it by themselves without the help of RSM. My goal is to work myself out of a job with my customers. I want them to be able to do this themselves. A self-sufficient customer is a happy customer. But I think a lot of times, you have to find the right people in the organization who can do it and follow through.

Think about Microsoft Excel. I go back to COVID — do you remember the UK? Their COVID tracking system broke one day. Somebody figured out they had used up the last row in their Excel spreadsheet for COVID tracking. The reason it broke was because they were using Excel as their backend database. I just remember laughing because the number they hit was the largest number of rows you can have in 32-bit Excel before it falls apart. I just about died laughing.

Samuel:
No, that’s—

Ben Vollmer:
Some of this is about giving customers — your end users — the right tools. How many times have you seen Excel misused? Because it’s the only tool someone has. IT only gives you Excel, so guess what you do? You do bad things in Excel.

So some of this is IT enablement and some of it is end-user enablement — making sure the right things happen at the right times. I’m really kind of excited by that.

Samuel:
I am as well. It’s really changing the whole way we’re working. I was a consultant for most of my career, and I would have loved to have all these tools available to me. But like you mentioned, delivering a project would have been completely different. You just mentioned that in some cases you’re more about enabling the customer — but what’s the biggest misconception you see companies having around implementing AI agents or implementing AI in general?

Ben Vollmer:
I’ll tell you — my dad’s an architect, OK? And I told my dad he had no reason to worry. A) He’s retired, and B) I would never take his job. He asked why not, and I said, “Because what you do is based on the questions you ask. Your questions come from experience. You know how to ask the right question to drive the right decision.”

Most people aren’t good at asking questions. And when you think about what you’re doing with prompt engineering, it’s really just asking the right question.

So I think the biggest misconception — actually, I was on a call yesterday with someone — they said, “I thought the shorter my prompt was, the better my response would be.” And I said, “No, no, no — it’s the exact opposite. Write your freaking War and Peace in your prompt and see what happens.”

I also think people confuse the hallucinations of public AI versus private AI. Like, think about it — Microsoft 365 Copilot is not the same as the Bing Copilot. The Bing Copilot reads the public internet for its data — so who knows what you’ll get in return — whereas M365 Copilot is much more grounded in the Microsoft Graph than public data.

And then when you have Copilot Studio — I just built an agent the other day that—

Samuel:
Mm.

Ben Vollmer:
—is grounded strictly in two documents. “Do not look up anything else; use these two documents for your knowledge.” So I think people’s misconception is that when they expose it, it’s going to look everywhere at everything — and narrowing it down is actually pretty hard to do.

Samuel:
I totally agree with that. That’s what I’m seeing as well — customers reaching out saying, “Hey, I have these 10,000 documents in a SharePoint folder. I want to plug an agent on top of it, and I’m expecting it to do the job of the whole department.” OK, that’s not how it works, unfortunately.

Ben Vollmer:
But I think helping them see — look, we can make your department better. We can get you better quality answers. I’ve been playing with Copilot Studio Lite just for fun. We have it enabled internally — we use D365, we use the whole M365 suite here at RSM.

I’ve built a couple of light agents, and I’m impressed — again, because I have a Copilot Studio background. It’s funny how far you can get with the right prompt. I think the misconception people have is they see it running out of the box and wonder, “Why doesn’t this work the way I want it to?”

Again, it’s like when you move into a new house — what’s the first thing you do? You put up pictures, mount your TV, set up your stereo — you make it yours. People see AI and go, “Well, this doesn’t work the way I want.” Well, G-I-G-O — garbage in, garbage out. Let’s put good stuff in and see what happens.

Samuel:
I like that example — the customer who thought shorter prompts would be better. And to your analogy of going from maps to GPS, it’s the same. We all learned to use Google by keeping searches short — “use fewer keywords to get better results.” Now it’s the opposite. I use full sentences, give lots of context, explain what I’m trying to achieve. But we’ve been taught for 20, 25 years to only use keywords — and now we have to relearn, like going from the map to the GPS.

Ben Vollmer:
Exactly. When ChatGPT first came out, the ability to personalize it was kind of funny. A buddy of mine didn’t personalize his at all, and I did — we compared results, and they couldn’t have been more different — like Venus and Mars. That personalization made all the difference. So personalize it. Make it yours.

Samuel:
If you had to pick one use case — you said you’ve started playing with Copilot Studio, building all those agents — there are so many that customers don’t even know where to start. So if you had to pick one use case that’s underrated today but will have huge potential in the next 12 months, what would it be? You mentioned OCR and computer vision — which one do you think is the most powerful?

Ben Vollmer:
I think it depends on where you work, Sam. If I’m a desk worker — document summarization, generation, document information, and analytics is my number one use case.

Samuel:
Yeah.

Ben Vollmer:
If I’m in the field — a field service tech, oil and gas worker, or any deskless worker — then I think computer vision and SLMs are going to be what drives progress over the next 12 to 18 months.

Samuel:
It’s mostly about acting on documentation better or faster than a human being would.

Ben Vollmer:
Yeah, but it’s like — we as humans only have a small number of connections. Think about how many people are in your frequent contacts — your spouse, your kids, your siblings, your parents, maybe a few close friends. That’s, what, 10 or 12 people? Even at work, your team might feel big, but it’s probably under 100 people.

If I look at computer vision — if I can take a picture of anything and ask, “What’s wrong with this object?” — and it can look across 2,000 or 3,000 technicians, all their work orders and photos, and come back with, “Hey, Sam saw this last week and fixed it by doing that,” — that’s harnessing the power of the entire organization to make me better as an individual.

Samuel:
Ben, we’re almost at the end of our time together. I always end with two questions — one about a personal tip, and one about your vision. This first one’s a tough one. What’s your number one productivity tip using AI in your own work — one you can’t live without?

Ben Vollmer:
I’ll tell you one I can’t live without. I made an agent in M365 Copilot called the Louis Litt Agent. Ever watch the TV show Suits? I created an agent named after Louis Litt. It includes the RSM 5Cs and a touch of sarcasm — because I’m a little sarcastic.

When I get an email where I know what I want to say but can’t really say it — my response is basically two words I can’t repeat on YouTube — I paste it into the Louis Litt Agent, and it turns it into something polished that I can actually send.

That’s my number one productivity tip. I created an agent that not only helps me but also makes me laugh every time I use it. I’d suggest doing something like that — it gives you a bit of serotonin, a little joy when you do it.

Samuel:
Can you share those instructions? I totally want to use that agent. That’s a great idea.

Ben Vollmer:
I don’t know what— It’s my favorite agent so far.

Samuel:
Last question: looking ahead 10 years, how do you think AI will change the way we live and work?

Ben Vollmer:
I really pray — from the bottom of my heart — that AI learns how to do laundry. I’ve got an AI vacuum cleaner that runs around my house every day, and I’m happy with that. But I wish AI would learn how to do my dishes and my laundry.

My 10-year goal is to have more AI infused into my life. But more importantly, I think of AI as an ability enhancer — something that enhances what I can already do. That’s the vision I want people to focus on: use AI as an ability enhancer.

Samuel:
Love it. And I’d love to have AI doing my laundry and dishes too.

Ben Vollmer:
You and me both.

Samuel:
Awesome. Thank you so much for your time, Ben. Today was fun — insightful and inspiring. Thanks a lot.

Ben Vollmer:
Thanks for having me on, Sam. It’s been a lot of fun.

Samuel:
Have a great rest of your day.

Ben Vollmer:
Thank you.