“We’re going to do ourselves a disservice … as an industry if we push people too fast into AI and we don’t give real solutions,” Insight SVP Stephen Moss tells CRN.

Artificial intelligence as a service. Managed AI. AI assurance.

Solution providers are at work sorting through new practices and new ways to not only capitalize on a new tech era, but bring customers a total package of AI software, infrastructure, consulting and service. In August, the International Data Corp. (IDC) put the AI market at about $235 billion in 2024 and should reach $631 billion by 2028.

Bringing customers a total AI package is essential for the emerging technology to take off with enterprises, Stephen Moss, senior vice president of managed services for Insight North America–a member of CRN’s 2024 MSP500–told CRN in an interview.

“We’re going to do ourselves a disservice … as an industry if we push people too fast into AI and we don’t give real solutions,” Moss said. “At that point in time, you’re selling stuff just to sell stuff.”

[RELATED: For Some Partners, AI Turbocharges Private Cloud Business]

AI as a Service

Rob Vatter, executive VP of platform services at Teaneck, N.J.-based Cognizant–No. 8 on CRN’s 2024 Solution Provider 500–told CRN in an interview that the AI market needs to settle questions around how to monetize this new technology before AIaaS becomes a reality.

Vatter said he does not mind vendors such as Microsoft experimenting with multiple AI business models in different products–Copilot for Security, for example, is consumption-based while Microsoft 365 Copilot is per-user.

Solution providers could see AI’s effects on their business show up as faster solution delivery, more revenue, better margins and potentially a mix of all that and more. “Instead of having labor as 70 percent of your cost base, it might turn out labor is 50 percent and your chip labor is the other 20 percent,” Vatter said.

Virtusa’s AI Assurance Offer

Surajit Bhattacharjee, SVP of technology and global lead for generative AI at Virtusa, told CRN that some might consider the Southborough, Mass.-based solution provider’s Helio offer AIaaS because it brings together consulting and engineering for AI products from conception to production.

However, Virtusa has purposely avoided the AIaaS label because customers might worry about a loss of control that can come with third-party as-a-service tools, Bhattacharjee said. The as-a-service label can also imply a level of indemnification that solution providers might not want to take on and might resonate more with smaller businesses looking for simple plug-in products.

“When it comes to large enterprises, which is where big AI will make the big bucks and service providers like us are going to remain valuable, there is the custom layer, really doing all the legwork upfront to make sure you don’t have the negative compounding of quality, being able to then ensure that this solution is good enough to go into production and subsequently keep monitoring it over weeks and months–that’s where we will play,” Bhattacharjee said.

Virtusa has described some of its AI service to customers as “discrete solution development,” covering AI proofs of concept, engineering work, pilots and bringing solutions to production, he said.

The solution provider has also tried out an offer for generative AI assurance, certifying GenAI products for safety and accuracy and continuously monitoring AI use cases in production. That continuous monitoring will prove especially useful in such a rapidly changing area of technology, he said.

Data curation is a term Virtusa has used for its services readying data for GenAI, especially with businesses that might have less of a handle on video, audio, images and unstructured data as opposed to structured data, Bhattacharjee said.

“Generative AI requires you to understand how to be intimate with this unstructured and rich content, how to analyze, process, semantically curate those types of content, and to be able to do that at scale,” Bhattacharjee said. “That’s the problem that most businesses are struggling with. So that’s going to be a big focus area for us in terms of services.”

Virtusa has also experimented with propositions with customers around lowered implementation cost in exchange for business based on outcomes, he said.

AI Platform As A Service

Neil Anderson, vice president of cloud, infrastructure and AI Solutions at Maryland Heights, Mo.-based company World Wide Technology, told CRN in an interview that business models that he sees emerging in the AI era include GPU-as-a-service, AI-platform-as-a-service machine learning operations (MLOps) and managed data streaming and source management.

Anderson said that the high cost of AI power consumption may have customers looking for as-a-service models instead of investing in their own data centers.

AI projects could offer solution providers post-deployment recurring data maintenance and model update work that could turn into an as-a-service offer from solution providers, he said.

“There are customers saying, ‘I’m not sure I’m able to keep up with that, maybe I will seek a firm that will help me,’” he said. “These things are living breathing animals that you just iterate on constantly. They’re never done, is what we’ve learned.”

It’s possible that AIaaS doesn’t become an individual offering, with AI instead applied to services already available to customers or just becoming part of existing vendor applications, multiple solution providers told CRN.

Vatter recalled telling one software vendor, “People pay a lot of money for your software already. … if you’re going to try to ask them for more money, this thing better be really, really good.”

“If you give it to them as part of the platform and it actually helps them use the platform, your churn rate on your software sales should reduce,” he said. “For me, if you can reduce your churn rate in software sales it would probably pay for having it as part of your software platform, as opposed to charging it as another” stock-keeping unit (SKU), which could upset customers.

Chandler, Ariz.-based Insight has considered using AIaaS or managed AI as a term with customers, Moss told CRN. But AI has also brought new power to Insight’s already-in-use “intelligent edge” technology offers. Insight also offers managed Nvidia platform and managed data services, which should entice AI customers. “With managed data, we can get to managed AI,” he said. “You can’t do managed AI and have no data.”

Anderson said that WWT has structured its AI offerings to meet a variety of maturity levels for customers, whether that’s in helping them define use cases or selling infrastructure. For WWT, it’s a matter of providing customers with the right solution that’s more important than more revenue.

“We can sell a lot of Nvidia infrastructure and we can make a lot of money on that,” he said. “But if the right solution for the customer is, well, you’re not going to have the sophistication to pull this off, let’s introduce you to a cloud-hosted solution.”

Rapid Development

Some solution providers are also thinking about the areas of the AI market where they might not have a presence. For WWT, some public sector AI markets might not fit well with its business. Virtusa might not advise customers around AI regulation and law, preferring to focus on engineering.

But for solution providers, no matter how AI changes the business, the goals for customers remains the same, Vatter said.

“The end goal is still efficiency, speed, accuracy, cost, satisfaction,” he said. “We just have better technology now to solve for them.”

WWT’s Anderson said that he predicts more off-the-shelf AI software coming, requiring less customization by users and coming with the potential for as-a-service.

Insight’s Moss said that he could see the AI market adding more configuration for customers, cutting down on the amount of customization work, perhaps giving way to AI-as-a-platform and AI-themed last-mile coding offerings.

No matter what offers and business models solution providers pursue, Moss said that his message to AI vendors is to involve partners in the emerging market and build an ecosystem around AI.

Putting effort into solution implementation instead of product development actually dilutes vendors’ earnings, meaning they should rely on solution provider partners, he said. If end customers don’t work with solution providers, they could get frustrated with AI products and “walk away” from the new technology.

“We all lose in that scenario,” he said. “Just selling AI for the sake of having sold AI in terms of hardware or maybe raw, unconfigured, unused software, that’s not going to do the industry any good.”