At Amazon’s (AMZN-Q) AI hub in Toronto, the tech giant and its partners say they’re selling a rarity in the world of artificial intelligence (AI) for business: fixed prices and guaranteed results. The strategy comes amid a slew of research on an “AI value gap” full of disappointing real world results at most companies attempting to use the technology.

Inside a ninth-floor downtown office, cartoon rocket ships and astronauts adorn white boards at what’s officially called the AWS Partner Innovation Hub. Amazon opened it last month. The space motif is meant to help tour groups of executives grasp various stages of AI deployment. These include “Liftoff, “Full Throttle,” and the "North Star." 

An Amazon Canada spokesperson says its clients include the federal government, Air Canada, Maple Leaf Sports & Entertainment, and the University of British Columbia. However, it keeps names of companies that have toured the hub private.

“We work with customers all the way from startups to full enterprises like the banks and insurance companies,” Patricia Nielsen, director of digital transformation and AI at AWS Canada, told TechNX. “Our partners can take their ideas and are able to move them into a formal plan, milestones, next steps and into production.”

Facing the "AI value gap"

Amazon knows most AI missions today fall short of their targets. Research conducted by the U.K.-based advisory firm Strand Partners for AWS shows that while 650,000 Canadian businesses now use AI, only 15 per cent have reached the stage where it genuinely changes how they operate.

And it’s not just Amazon highlighting this problem.

“While AI capabilities are accelerating, the ability to translate those capabilities into business value remains a challenge for many,” global accounting and consulting firm KPMG wrote in a recent report.  

“This gap between what leaders feel AI should achieve, and how prepared their teams and processes are to deliver on that promise, is referred to as the AI value gap.”

Deloitte, EY, and Boston Consulting Group also refer to similar “gaps” in recent research. 

According to IBM Canada, only 43 per cent of AI initiatives have delivered their expected return on investment over the past two years.

"Canadian organizations are still figuring out how to scale AI responsibly," IBM Canada vice-president and chief technology officer Manav Gupta stated in a news release last month.

“Closing that gap will be critical to realizing AI's full value and staying competitive.”

Touring Amazon's AI hub

Amazon says it tailors its 60-minute hub tours to specific companies and industries. Highlights include an interactive AI-powered Amazon vice president inside a life-size holographic display.

Then there’s a station where visitors can draw outlandish items on a touch screen, for example, a futuristic new car made of wood. The AI instantly spits out a stylized rendering, a suggested price, and detailed plans for a marketing campaign.

At another station, AI essentially listens to a visitor’s business challenges, and rapidly pitches a buffet of solutions based on preferences like cost or time. 

“It really helps you to see what’s possible, rather than just talking about it,” Nielsen said.

There’s also an optional four-hour workshop.

“We’re able to show them use cases and examples from other customers we’re working with around the world,” Nielsen said. “We’ve got thousands.”

“When we dig into the data, we’re seeing that their usage can be early stages, or not in use cases at all that are really moving the needle,” she added.

The hub’s partners include Montreal-based IT and consulting firm CGI and Dedicatted, a Germany-headquartered boutique consultancy. There’s also Elevata, an AI-native cloud consultancy formally based in the U.S., as well as Vancouver-based OpsGuru, which essentially helps companies clean up their data so tools like generative AI can go to work.

OpsGuru's experience with the hub

OpsGuru CEO Ryan Smyth sees “data fragmentation” as the biggest hurdle for companies with enterprise AI.

“(That’s) where they’ve got systems that are spread all over the place,” he told TechNX. “The data is not their own, and certainly not curated in a way that they can access it in a manner suitable for layering AI on top of it.”

OpsGuru says it’s the first in North America to offer fixed-fee pricing for AI-native professional services. That’s where AI agents do the core production work, and human experts act as architects, auditors and project managers.

“Speed is really changing the equation. Transformation in the past used to mean a monolithic project. These monolithic projects would typically charge on a time and materials basis, because they were very slow. What’s changed with AI is the speed at which we can deliver outcomes, and the way in which it can be more targeted,” Smyth said.

“With our experience and the tools that we have, we are able to provide a guarantee of an outcome for a customer.”