IBM Consulting’s Mohamad Ali Details ‘Human Plus Digital Labor’ AI Model

‘I can actually watch and look into the factory and see 600 projects right now, real time, being manufactured with human plus digital labor. It’s almost like looking into a car factory and seeing the humans and the robots build a car,’ says IBM Consulting Senior Vice President Mohamad Ali.

Mohamad Ali, senior vice president of IBM Consulting, is guiding the solution provider giant through agentic artificial intelligence adoption internally and with customers—a benefit, he says, not only to IBM but to the solution providers that partner with his organization.

“We probably have hundreds, thousands of subcontractors that we work through,” Ali told CRN in an interview. “When we do [some projects], we bring together 50-plus companies in order to execute that.”

And as much as IBM Consulting—No. 8 on CRN’s 2026 Solution Provider 500—is a resource for some IBM solution providers, Ali isn’t afraid to put his organization above the competing global IT firms.

“We are at least 18 months ahead of everybody else,” he said. “And the numbers that we see are showing that.”

[RELATED: IBM CEO Krishna: Q2 Weakness Due To Deal ‘Deferral, Not Destruction’]

IBM Consulting Combines Human Expertise With 6,000 AI Agents

Ali caught up with CRN during a period of growth and high expectations for this IBM segment.

In the 115-year-old Armonk, N.Y.-based technology giant’s most recent quarterly results—covering the three months ended June 30—the consulting division brought in $5.3 billion in revenue, up 1 percent ignoring foreign exchange. IBM Consulting saw signings grow 6 percent year on year, a second consecutive quarter of growth.

IBM Consulting’s profit margin also grew 160 basis points during the quarter, faster than IBM Software’s 110 basis points. IBM Consulting reported a gross profit margin of 28.9 percent for the quarter and segment profit margin of 12.1 percent for the quarter.

IBM expects IBM Consulting to accelerate in revenue growth to low- to mid-single digits for the year, according to the vendor.

Here’s more of what Ali had to say to CRN.

What should IBM partners know about you?

Not that long ago, I ran a software company where we served basically 500,000 SMBs. And it was all through channel partners. I think we had 20,000 VARs and resellers.

I’m on the other end of the spectrum now where our partners tend to be larger partners. But we also have a lot of smaller partners. And so there is quite an ecosystem around IBM Consulting. Maybe not dissimilar to other consulting companies.

We probably have hundreds, thousands, of subcontractors that we work through.

When we do [some projects], we bring together 50-plus companies in order to execute that. So there’s very much a set of channel partners that we’re highly dependent on. But we’re also a channel for the Microsofts and the Azures and the GCPs [Google Cloud Platform] and all that stuff.

I’m not a career consultant. I spent 30 years in software and products and so forth. And I was with IBM in the early days. [IBM CEO Arvind Krishna] and I were peers. We reported to the same guy for four years.

After I sold the last company that I was at to Blackstone, Arvind asked me to come back.

I said, ‘I’m not a consultant.’ He said, ‘You don’t understand. Consulting has become a human plus digital labor business. And digital labor is lots of bits of software.’

That’s what I’ve been doing for two and a half, almost three, years—creating a human plus digital labor business.

How does IBM Consulting compare to the competition?

We are at least 18 months ahead of everybody else. And the numbers that we see are showing that.

We have an over $10 billion book of business that is delivered with human plus digital labor. And this isn’t theoretical.

I can actually watch and look into the factory and see 600 projects right now, real time, being manufactured with human plus digital labor. It’s almost like looking into a car factory and seeing the humans and the robots build a car.

By doing what we’re doing, it now allows us to reach smaller customers that we didn’t reach before. It allows us to really participate in the midmarket because of what we can do with much smaller teams. We could be much more nimble.

And that was actually not one of the original objectives. But that’s starting to happen.

We recently had a deal in Europe. And it was a pharmaceutical company, and we were bidding against what you would consider the low-cost, offshore providers.

And because our solution included this powerful combination of human plus digital labor, oddly enough, we were the low-cost provider. You don’t really think of IBM as the low-cost provider. But in this case, we were both high quality and low cost.

You can see that in some of the projects we’re doing now. And if we could do that uniformly and consistently, we would be a consulting company unlike anything we have ever had at IBM.

How has IBM Consulting evolved for the AI era?

We built this environment where we have now these 6,000 digital workers.

We have 150,000 consultants, and almost three years ago when we started, every one of our competitors went out and bought lots of AI licenses, handed them to their employees and said, ‘Go be productive. Here are 100,000 [Microsoft] Copilot licenses. Here are 100,000 OpenAI licenses, whatever. Go be productive.’ And it just didn’t work.

Our approach was, we built this uber harness. We call it IBM Consulting Advantage. And below we’re going to put all the AI you could possibly want. Any stack. Goes underneath.

We’re going to abstract the digital workers on top. And a digital worker is anything that calls an LLM. So anything that’s nondeterministic.

We’re going to focus on building these use-specific bits of AI. And when we first did this, we ended up with like 50,000 of these digital workers, which is not very useful.

I hired a product management team from software. And we triaged it, QA’d it.

We've now gotten it down to 6,000 [digital workers]. That’s in production.

We put them into these packs. So there are 155 packs. If you’re going to manufacture an Adobe project—like a front end for a bank, let’s say—then you get this pack of say 20, 30 digital workers. And you’ll get a pack of 10 human beings. And we’ll put them together.

And you can watch them in a chatroom as they come up with the user stories, as they build the code, as they test the code. And this could go on for hours and days or weeks or months. And at the end it’s done.

Then we organized it by [division]. Our divisions are technology transformation, business transformation for applications like SAP, hybrid cloud—which is AWS, GCP [and other cloud] migrations—AMS [application management services], BPO [business process outsourcing] and cybersecurity.

[IBM’s Autonomous Threat Operations Machine, or ATOM, for example, has] 30 digital workers. They do threat investigations.

What this means is that if a threat comes into a SOC [Security Operations Center] and it’s significant enough, we might ask [a human] to investigate it.

[That human could] have taken 45 minutes. [ATOM can do it] in two minutes. And then [the human] will spend 13 minutes checking it.

Last month we did 70,000 investigations. And most of them, when [the human] checked them, they were fine.

Can you give me a sense of scale for the agentified IBM Consulting segment?

IBM Consulting Advantage uses about a half a trillion tokens per month. This thing is running at massive scale. So that’s what we build to hold the digital workers and to deploy them into our processes.

We have over 2,000 large projects that we do every year. And you can see about 600 of them now are being manufactured with human plus digital labor.

We have something called a model router that optimizes the use. So I know that on average last month it cost me $2.58 per million tokens. Very few other organizations can tell you that.

In the old days, if you wanted to re-engineer a process [like] an accounts table process, some consultant would come with stickies and they would put it on a board and they’d map out the process. And then you’d work to figure out how to re-engineer it.

We don’t have to do that anymore. You could watch this thing run on the weekend. … I don’t think any of our competitors have that and have that running at this massive scale.

You can see that in their numbers. None of their profits are really increasing.

If you look at [IBM] Consulting, every quarter it is going up. … [It’s up] more than 20 percent in 24 months. Almost every quarter, it’s going up. And it’s because we’re driving this kind of efficiency.

How about re-engineering client processes?

We’ve mapped out these human plus digital blueprints for banking, for telco, for corporate functions.

So if you wanted to re-engineer, let’s say accounts payable, you need three things.

You need to know what the existing process looks like. Some of our clients have that. That’s a place we might use a channel partner who might go in and do the mapping for us.

The second thing that we need is an idealized view of what an accounts payable process is. And we’ve done 30 years of this. So we basically codified all of that into the system.

We created this blueprint for reimagining this as human plus digital. Here is the accounts payable process. It’s got all these atomic processes, subprocesses. This one we say we use deterministic code [for example]. This one we say we’re going to use AI. And if we’re going to use AI, then how does the human in the loop fit in? There’s this whole blueprint here about how we’re going to re-engineer it.

[For one client], they gave us their processes. We had these agents go through it, figure out the lens, the technology stack, the pain points. They also figured out what’s missing from their process.

You don’t have all this stuff documented, but we can still work with it. So then it starts re-engineering their process. We have a [workflow analysis tool] called Procedure Eater. You can watch it run on the weekends.

It analyzes their standard operating procedures, their SOPs.

Then it says, ‘Well, you have 18 atomic processes. And if I re-engineer them with human plus digital, nine of them I can do really well. One of them I can’t do so well. And then at the end of that, here’s the FTE [full-time equivalent employees] reduction.’

And then it goes and it builds a whole target operating model [which shows how an organization structures its people, processes, technology and governance to execute its corporate strategy].

You get a PowerPoint because that’s what you could submit to your board or to your management team.

Let’s say you decide you’re going to build it. Then this Process Studio actually goes out and builds it. It starts spec’ing out all the agents that you need.

It says, ‘Here are all the agents that you need to re-engineer this thing. Here are the specs.’

And then you can take the spec and put it into [OpenAI’s] Codex or [Anthropic’s] Claude Code, etc.

These are all the specs for building the agents. Here are all the connectors, the MCP [Model Context Protocol] servers. You could spec these as well. You basically take all this output, put it into a coder, and guess what? You have the agents.

It is just miraculous. [Before], this would take us months or years. Sometimes you would just not even bother because it was so complicated. This is the new IBM Consulting.

And you could see how with what we’re doing here, the channel is even more important to us.

There are lots of pieces where we’re going to depend on the channel, too. And then we can also start reaching clients that are not just the mega clients, but also the midsize clients.

How about the economic model for these engagements?

For these projects, we are shifting more and more to fixed-price projects because we have so much confidence now in our ability to do them. That is our preferred mode.

We also like outcome-based pricing. We often propose that. There are clients who love that. There are clients who just don’t know how to consume that yet because the procurement department wasn’t built that way.

But I would say that that fixed-price and outcome-based are two of the directions that we are moving more and more into.

Clients have come to us and said, ‘Hey, that’s great. Can we buy some of your digital workers?’

We have not really figured out the licensing and pricing mechanism for that, but that is the next frontier.

We’ve done a few of these deals where we’ve licensed the digital workers and this uber harness and then put it on top of other AI stacks.

This big deal that we announced with GCP recently, it is really all about taking our digital workers and our IBM Consulting Advantage and [we] put it on top of GCP as an important AI stack.

We’ve actually partnered with Pearson. And they credentialize human beings.

We’re doing the same thing for these digital workers now.

The digital workers in Consulting Advantage will periodically reach out to the Pearson [credentialing tool] and get their badges that way. So we’re not certifying [the agents].

Same way that if you want to buy on a rate card basis time and materials, you’re buying X engineer or Y engineer for a certain rate, should you be buying a digital worker for a certain rate as well?

We’re just starting to experiment with that. But what we’re doing in our Consulting business is we are using these digital workers to solution and to deliver, and we’re moving to more fixed price.

Will this differentiation for IBM Consulting last as we move further into the AI era and more consulting firms leverage AI like this, if they can?

The encapsulation of skills and knowledge is going to be the key to successful AI implementation.

We have got these 6,000 digital workers. And if you think about it, they are an encapsulation of our knowledge. And that’s hard to come by.

Our competitors haven’t bothered to do that. Maybe they will at some point.

We’ve encapsulated all this process knowledge of how processes work—which is the core of what Consulting does—into these tools like Process Studio. We’re turning these things into bits of AI.

The third area is what we call skills. If you look at many of these tools out there, they ask you to build the skills natively in their environment.

We don’t do it like that. We build the skills in Consulting Advantage and then download the skills into somebody’s AI tool. So we protect that IP centrally.

Pick somebody’s popular AI tool. And if you turn it on, look at the screen. There’s the thing that says ‘skills.’ We don’t leave all the skills there. The skills actually reside centrally in [IBM Consulting Advantage].

And when you’re ready to use that tool, it downloads there, and then it disappears. So this central mechanism for protecting this IP that these human beings continuously create I think is the moat.

And the bigger that pool of IP, the harder it is for another consulting company to catch up. Which then makes us an incredible channel partner to the OpenAIs, the Anthropics, the [Google] Geminis, as well as the [Microsoft] Azure AI Foundry stack, the AWS Bedrock stack, etc. This is the moat we’re building.

Is IBM Consulting a resource for channel partners?

When we win these projects, we have to go execute them. And we have a whole bunch—I mean, hundreds, if not thousands, of channel partners that want to execute them.

And we do enable them with tools to execute. Because if we’re executing one way, they’re executing another way, that’s a problem. We do bring them into this fold.

We do have quite a large channel as a contractor.

What are your thoughts around growing interest in forward-deployed engineers, FDEs, from technology vendors?

Part of why IBM has been successful for many, many, many decades is that we don’t just sell you products. We bring some of the smartest people on the planet to implement it.

Almost every single day, transactions that you make go through some certified IBM system. These things are rock-solid. And they don’t ever go down.

We have some of the best engineers. And for a long time, we’ve included our IBM fellows, our IBM distinguished engineers as part of that client engagement team.

What we have done in consulting is we’ve created these units called FDUs—forward deployed units. Because a single engineer is not as helpful as the unit. … A small unit is about three. And a large unit is about eight.

You need a senior person to actually architect how we’re going to engineer it, how we’re going to use the human plus digital, and then probably two junior people who can actually go run the Process Studio, get the agents out, use the code builders to build the agents, qualify them.

That’s an FDU. So we’ve been providing these FDUs now for several quarters to our clients as part of our engagement with them. … When we have a digital worker, [the client also gets] one small FDU along with it to install it.

But if you want to build 12 of these digital workers for yourself, you’re going to want some of our FDUs along to build them.

And so we’ve also entered into partnership with AWS and some of the other firms who have FDE practices that they’re building. But how are they doing that? Part of how they’re doing it is through FDUs that we bring to the table, and they’re effectively a channel for us.

They go out to a client. They win some big deal. They need some FDEs or FDUs. We supply that. We do the AWS work.

Once we’re engaged, we say, ‘Well, we also have some digital workers that we’ve built for these other purposes. Would they be useful?’ And you can see how the cycle goes.

So how do you view IBM Consulting’s competitive landscape today?

When we go to a client, we have to bring the best technologies. And oftentimes, the best technology is an IBM product like Red Hat.

We also bring Microsoft technologies. We have 33,000 Microsoft-certified professionals within IBM Consulting.

We just signed a big deal with Google. We have a whole Google practice with many people in it. We have a large AWS practice. We have a large SAP practice, Oracle practice, etc.

IBM Consulting is a human plus digital labor business. This is a huge differentiator. And I don’t think our competitors are close behind. They’re at least 18 months away. And this is what our clients are looking for.

We’re operationalizing at scale. We’ve got hundreds of projects live that we could look at, and it’s driving our margin expansion. … If you were to look at our top four competitors and try to look at their profit over the last eight quarters, none of them are really growing their profit.

We are growing our profit like every quarter. So you look at that and you say, ‘Am I really going to trust those companies to come and re-engineer my beer company?’—or whatever company I have—to be more efficient when they can’t actually get their own companies to be more efficient? And that is a differentiator.

IBM Consulting has always been known for high quality—if you want to get a project done and done right, you go to IBM Consulting. … Where we’re going next is high quality at affordable costs.

Everyone a decade and a half ago did labor arbitrage. They put people offshore. The game is different now. It’s technology arbitrage.

That’s what we need to lead in. And if we can lead in that, we can have affordable solutions that are high quality.

Does your granular view into projects reflect your leadership style?

Yes, I do get into the details. … Every couple of weeks, I actually have somebody who comes to my office. And he reverse-mentors me, and we write code together.

I don’t know how many CEOs of consulting companies do that.

I'm on that [IBM Consulting Advantage] dashboard a dozen times a day. … I will go into projects, and I will see the quality of the digital workers they’re using.

We have to build a product that is better than anybody else’s products. But then we’ve got to operationalize it, and to operationalize it requires attention to detail and focus.

Has this flurry of attention around IT security spurred by Anthropic Mythos impacted IBM Consulting?

Security, absolutely. … In Consulting, that is a double-digit growth in business for us.

We’re leading in certain areas. I touched on the Autonomous Threat Operations Machine, which is the 30 digital workers that do threat investigations. We did that before anybody else that I know in our consulting competitor sphere.

Just like you’re being attacked in an autonomous way, we now have a fleet of digital workers that we deploy on top of our partners’ platforms, like Palo Alto [Networks’] security platform. Like Microsoft’s security platform.

But these digital workers sit on top of this stack. And they are defending you at the speed that the AI is attacking you. That’s been a big initiative for us.

We will use various AI models to effectively scan the environment for vulnerabilities. We can do that right off of that Consulting Advantage environment.

And then you have the patch. We have to do the front-end assessment and the scanning, and that is going well for us. And then on the back end, the patching. And there are two types of patching.

There’s open-source patching. And we have Project Lightwell. So we work with our Red Hat team. We bring those patches to bear.

And then there’s non-open-source-patching. And we have a set of tools to do that patching as well.

Quantum safe, that’s also another area that IBM Consulting leads. I don’t think any of our competitors have got the capability that we have.

On the security side, there’s a lot of opportunity for consulting companies, but they have to come in with better tools, a better offering.

That’s what you see in the IBM autonomous security agents that we have. That’s what you see in the quantum safe capabilities we have.

How is AI affecting consulting economics?

If you look at our business in consulting, that part where we are doing AI work is growing substantially faster than the rest of the business.

But what we’re also seeing is that there’s a part of the traditional business that’s compressing.

If you are a car company and used to pay $100 million [for example] to do application management services, now you want that for $70 million because you just assume AI should do that.

And so that part of the business is compressing for us. It’s compressing for our competitors. And that’s why it’s so important for us to grow this new business, this AI business, this security business as fast as we can.