Want to incorporate AI into your business? Who you gonna call?

NOTE: See also main story, “The ‘best-kept-secret’ in Canada’s innovation ecosystem is seeing increasing demand for its services.”

Canadian companies looking to incorporate artificial intelligence into their business but lacking the know-how and resources to do so have a college-affiliated Technology Access Centre (TAC) that can help.

The AI Hub, a TAC at Durham College in Oshawa, Ontario, offers companies a full suite of services to help them with the entire range of designing, adopting and deploying AI. This includes identifying what the AI use case is to what type of technology is suitable, helping execute an AI roadmap, and then integrating AI into their operations.

“We aim to be able to take the company from idea to execution,” Natalie Arthurs, (photo at right), senior manager of the AI Hub, said in an interview.

“Companies can be anywhere on that continuum and we can meet them where they are. Our suite of services is lined up to take a company from concept to reality.”

The AI Hub can help companies deploy the full scope of AI technologies, including traditional AI like machine learning and computer vision, to newer applications such as generative AI.

Arthurs said the AI Hub mainly services two types of businesses. One group is mostly  medium-sized businesses with more than 10 employees that are generating sufficient revenue to invest in new technology, but may not have the expertise or resources to build out an AI development team.

“They partner with us to become like an extension of their organization and get them over the hump on developing their AI strategy and executing it,” Arthurs said.

These companies tend to come from established businesses that are data-rich, because these are the firms that really need AI to take full advantage of their operational data, she said.

For example, companies might work with the AI Hub to use all of their operational data to build a virtual agent that can either answer custom questions for their clients, like a chatbot, or create a virtual agent that can help companies perform tasks and automate tasks that are routine.

These companies typically work with the AI Hub in “sprints’” lasting four weeks to four months.

The top five sectors the AI Hub works with are health and wellness, financial services, energy, logistics, and digital marketing, Arthurs said.

The other type of businesses the AI Hub serves are AI or tech companies that are on the smaller end of SMEs and may be pre-revenue.

“They’re looking for a long-term R&D partner to help them go through the commercialization lifecycle and overcome the range of barriers related to commercialization, including intellectual property, funding, development expertise, access to cutting-edge research and access to compute power,” Arthurs said.

These companies typically work with the AI Hub for six months or longer and in phases. The more research-intensive or go-to-market companies either leverage available government innovation funding or are in more of a bootstrapping stage.

AI Hub offers four “buckets” of services

The AI Hub offers four main “buckets” of services, Arthurs said.

The first bucket is strategic advisory or consulting services, which includes working with a company to develop their detailed AI roadmap that will take them from idea to execution.

The roadmap identifies everything the companies will need to move along that continuum, from the data they’ll need and what they’ll need to do with it, to the compute infrastructure and AI architecture required to make their concept proprietary and deployable.

The AI Hub’s next bucket of services is “data and AI services.” This essentially entails helping the company develop a custom AI solution for its operation – everything from building and testing to optimizing and deploying.

“One of the most popular solutions that we’ll build are custom virtual agents, beyond just answering text-based questions,” Arthurs said.

The AI Hub also helps data-rich companies build AI applications like dashboards, which can help the firms drive insights out of their operational data.

The AI Hub’s third pillar of services is professional training, both for technical and non-technical audiences.

On the technical side, the AI Hub will provide custom training for companies that, for example, have built an AI solution with the AI Hub and wants to train the companies’ workforce on how to use it.

On the non-technical side, the AI Hub has, in partnership with the Council of Canadian Innovators, launched a new AI Guide online training program for non-technical professionals in Canada’s business and information sectors, including leaders in human resources, sales, operations and public affairs.

The fourth bucket of services the AI Hub offers focuses on research-intensive projects, mainly with those clients looking for a long-term R&D partner. This entails helping with prototyping, developing proof-of-concepts and minimum viable products.

Some examples of these types of projects include:

• A mid-sized company in the electricity industry that partnered with the AI Hub to develop a machine-learning powered solution to significantly reduce the time required to process nuclear reactor inspection data. The AI software enabled the company’s team to access results faster and more accurately.
• A startup technology company in the corporate law sector that partnered with the AI Hub to create an AI tool that automates data analysis and document generation. The technology allows law firms and corporations to manage their records more efficiently and accurately.
• A Toronto-based edtech firm that delivers assessment and learning solutions tailored to meet the needs of students, educators and administrators partnered with the AI Hub to address growing concerns around academic integrity. The company was able to integrate machine learning, predictive analysis and advanced data operations into their e-assessment platform, enabling them to identify suspicious student behavior more effectively while addressing challenges such as algorithmic bias and privacy concerns.

Enhanced benefits from AI Hub’s college affiliation

The AI Hub being a TAC located at Durham College offers companies enhanced benefits, in addition to the equipment and expertise of the AI Hub itself.

Within the AI Hub, companies have dedicated access to compute power with GPU accelerators (GPU, or graphics processing units, are the core of AI development), multiple high-performance computer workstations, robotics to help with development, AI and machine learning toolkits, and facilities to accommodate training needs for small and large groups.

The AI Hub is part of Durham College’s Office of Research Services, Innovation and Entrepreneurship, and is co-located on campus in the Centre for Innovation and Research with the college’s two other immersive technology hubs, focused on cybersecurity and mixed reality capture (the whole suite of augmented and virtual reality).

Companies working with the AI Hub also have access to these two other tech hubs, Arthurs said. “We bring the full weight of those resources to every project and every client, and that is an incredible advantage, particularly as these worlds start to collide.”

Cybersecurity, for example, is the foundation to keeping everything, including security-sensitive data, safe and secure in the world of AI development.

“The weight of those three facilities really broadens the type of clients that we can take on, and ultimately the impact and the results that we can deliver for those clients,” Arthurs said.

The AI Hub also has made significant investments in the compute infrastructure required to run, train and refine complex AI models, and has the technicians who know how to optimize this compute power.

Small and medium-sized companies wouldn’t be able to afford such powerful computing resources, which might prevent them from taking any further steps in developing their AI solution, Arthurs noted.

 In addition to the AI Hub transferring all intellectual property to client companies, the AI Hub also has the in-house expertise to provide companies with advice on mobilizing and protecting their IP.

Durham College participates in the pilot program in Ontario that is focused on generating, protecting and leveraging Canadian-generated IP, Arthurs said. “Our clients also have access to that advice and potentially funding to protect their IP.”

All projects and initiatives the AI Hub works on are generated by industry. The AI Hub offers seasoned subject matter experts, often working in industry, to help lead projects.

The projects also are supported by a team of college students and recent graduates who typically have gone through Durham College’s post-graduate program.

“They’ve worked with the company, they know their data, they know the problems they’re trying to solve,” Arthurs said. “And at the end of the project, often they’ll get hired on by the company.”

All of the AI Hub’s services can be delivered online, so the hub serves a mix of clients in the Oshawa-Whitby region as well as companies across Canada.

Since 2017, the AI Hub has partnered with more than 150 companies and leveraged over $5 million in investment directly into industry-driven projects, Arthurs said. The hub currently supports about 50 companies per year – and demand is growing.

AI is a cross-sectoral platform, so the AI Hub works with other TACs and sectors looking to adopt and deploy AI solutions, such as agriculture, biochemical and other companies.

What is the demand for the AI Hub’s services? Arthurs said the AI Hub last year opened a call for proposals in partnership with the National Research Council’s Industrial Research Assistance Program. When the call for proposals was closed, there were 30 companies on the waiting list.

This year’s call just went out last week, and the day after the call the AI Hub had already received 20 applications from companies.

“People need the AI Hub’s services for sure,” Arthurs said. “The demand is there, 100 percent.”

R$